Upload
others
View
5
Download
0
Embed Size (px)
Citation preview
Brigham Young University Brigham Young University
BYU ScholarsArchive BYU ScholarsArchive
Theses and Dissertations
2016-05-01
Placental ‘��Omics’�� Study to Understand the Pathogenesis Placental ‘��Omics’�� Study to Understand the Pathogenesis
of Preeclampsia of Preeclampsia
Komal Kedia Brigham Young University - Provo
Follow this and additional works at: https://scholarsarchive.byu.edu/etd
Part of the Biochemistry Commons, and the Chemistry Commons
BYU ScholarsArchive Citation BYU ScholarsArchive Citation Kedia, Komal, "Placental ‘��Omics’�� Study to Understand the Pathogenesis of Preeclampsia" (2016). Theses and Dissertations. 5876. https://scholarsarchive.byu.edu/etd/5876
This Dissertation is brought to you for free and open access by BYU ScholarsArchive. It has been accepted for inclusion in Theses and Dissertations by an authorized administrator of BYU ScholarsArchive. For more information, please contact [email protected], [email protected].
Placental ‘Omics’ Study to Understand the Pathogenesis of Preeclampsia
Komal Kedia
A dissertation submitted to the faculty of Brigham Young University
in partial fulfillment of the requirements for the degree of
Doctor of Philosophy
Steven W. Graves, Chair Steven R. Goates
Mathew R. Linford Paul B. Farnsworth Matt A. Peterson
Department of Chemistry and Biochemistry
Brigham Young University
May 2016
Copyright © 2016 Komal Kedia
All Rights Reserved
ABSTRACT
Placental ‘Omics’ Study to Understand the Pathogenesis of Preeclampsia
Komal Kedia Department of Chemistry and Biochemistry, BYU
Doctor of Philosophy Preeclampsia (PE) is a potentially fatal complication of pregnancy characterized by an in-
crease in blood pressure (>140/90 mmHg) and proteinuria (>300 mg/24 hrs), often accompanied by edema. Symptoms of PE start after 20 weeks of gestation. If PE remains untreated, it can lead to eclampsia, grand-mal seizures responsible for most fatalities. PE is believed to affect 2-10% of pregnancies worldwide, and claims the lives of over 75,000 mothers and 500,000 newborns yearly.
No therapeutic agents have been developed to prevent or cure PE. Part of the reason for
this is the absence of a complete understanding of the pathogenesis of this disease. PE has long been regarded as a “disease of theories”, and the pathophysiology of PE continues to be the sub-ject of debate. Nonetheless, several abnormalities have been observed to precede established, clinical PE and have in turn been proposed to be involved in the causation of this disease, all with involvement of the mother’s placenta as a central feature. Removal of placenta is the only cure for PE and results in a rapid resolution of the symptoms. Thus, the placenta remains an or-gan of substantial interest and many research groups have attempted to identify abnormal placen-tal features occurring in PE. None of these studies have focused on less abundant, low molecular weight (LMW) biomolecules, which play important roles in the pathophysiology of many dis-eases.
There are a number of alterations that are believed to affect the placenta and contribute to
the pathogenesis of PE. The most widely accepted ones include hypoxia, oxidative stress, and an increase of pro-inflammatory mediators in the mother’s placenta. The goal of my initial study was to identify which of these hypothesized causative pathways has a significance in the etiology of this syndrome as well as to investigate which less abundant, low molecular weight biomolecules change in response to these abnormalities. For this purpose, we first adapted and optimized a pre-viously developed methodology that studied LMW biomolecules in tissue specimens to study pla-cental biomolecules. This approach involved a tissue homogenization step followed by protein depletion using acetonitrile. We compared two regions of human placenta: the chorionic plate and the basal plate to find differences in the LMW fraction. We discovered 16 species with statistically significant differences between the two sides, and identified 12 of them using tandem mass spec-trometry.
In the second study we collected normal human term placentas from elective C-section de-
liveries and exposed explants to each of the above-mentioned provocative agents or stress condi-tions for 48 hrs. Other explants without any stressors were cultured in parallel for the same amount of time. The processing of explants was divided into five steps: 1) explant culture; 2) tissue ho-mogenization; 3) acetonitrile precipitation to remove high abundance, high molecular weight pro-teins; 4) injection of the protein-depleted specimen into a capillary liquid chromatography–mass
spectrometer; 5) analysis of MS data to identify quantitative differences between cases (stressed explants) and controls (normal explants). In total, we observed 146 molecules changed in abun-dance between the treated explants and the controls with 75 of these molecules changed in response to hypoxic treatment, 23 changed due to hypoxia-reoxygenation, a process generating reactive oxygen species, and 48 changed due to tumor necrosis factor–alpha (TNFα), a pro-inflammatory cytokine. We were successful in identifying 45% of all these molecules by tandem MS. Statistical modeling that applied LASSO analysis allowed for the development of a model that used 16 of the 146 differentially expressed biomolecules to accurately classify and differentiate each of the 4 stressed conditions.
In my third study, I then submitted actual preeclamptic and non-diseased placental tissue to
our established homogenization and acetonitrile precipitation protocol to see if any of the differ-ences in LMW biomolecules produced under stress conditions in normal placenta were recapitu-lated in actual diseased placenta. In a preliminary statistical analysis, 8 of the original 146 differ-entially expressed species, displayed significant or near significant changes in the actual disease placenta. After applying two stringent statistical tests that eliminated any potential influence of gestational age, four out of the 146 biomarkers previously studied, continued to be differentially expressed in both stringent analyses. Of the four, 1 biomarker (m/z 649.49 (+1)) showed an in-creased abundance in hypoxic placental explants as well as in PE placenta; 2 (461.06 (+1), 476.24 (+1)) were increased in response to TNFα-exposed placental explants and in these PE placentas and 1 (426.35 (+1)) increased in response to hypoxia-reoxygenation-treated placental explants was also increased in PE placenta. We have chemically characterized 2 of the 4 biomarkers. One was a phospholipid (m/z 476.24) while the other was an acylcarnitine (m/z 426.35). This suggests that features of PE appear to arise from the predicted early abnormalities that affect the placenta.
In conclusion, I was successful in developing an ‘omics’ approach to study less abundant,
low molecular weight biomolecules in human placenta as well as investigate which biomarkers show differential expression in human placenta when exposed to proposed abnormalities of PE and have data to suggest that these same responses are present in PE placenta .
Keywords: Tissue proteomics, top-down, capillary liquid chromatography–MS, low-molecular-weight proteins, placenta, peptidomics, lipids, metabolites
ACKNOWLEDGMENTS
I would like to express sincere gratitude to my advisor Dr. Steven W. Graves for providing
me this wonderful opportunity to participate and perform research in his lab. His unwavering
support, guidance and encouragement helped me become a better scientist, and made my experi-
ence as a graduate student very pleasant. I am also grateful to my research collaborator Dr. Craig
D. Thulin for his encouragement, advice and extremely valuable insights into the field of prote-
omics. I would like to thank Bruce Jackson, the manager of the mass spectrometry center for his
constant guidance and assistance with instrument related concerns. I will always be very appre-
ciative to the department of chemistry and biochemistry, BYU for providing me with excellent
opportunities to excel as a graduate student. In particular, I would like to thank our assistant
graduate coordinator Janet Fonoimoana, whom I saw a friend in, for her help, kindness and con-
stant encouragement throughout the program. Graduate and undergraduate students in particular
Dr. Moana Hopoate, Dr. Tanielle Alvarez, Caitlin Nichols, Stephen Smith, Taryn Prestwich and
Andrew Wright are acknowledged for their direct contribution to this work. I received a substan-
tial amount of help in statistical analysis from Dr. Dennis Tolley and his graduate student Justin
Barnes which I greatly appreciate. I would also like to express my appreciation to my co-workers
Dr. Dipti Shah, Swati Anand and Ying Ding for creating a very fun-filled and positive atmos-
phere in the lab.
I also want to give credit to extremely talented professors in the chemistry department, es-
pecially my graduate committee members Dr. Steven Graves, Dr. Steven Goates, Dr. Paul Farns-
worth, Dr. Matt Linford and Dr. Matt Peterson for their encouragement, valuable insights and re-
views which helped me become better graduate student and researcher.
I am indebted to my mother Renu Kedia and father Anil Kedia for their unconditional love
and faith in me. Amidst daily struggles, they created an environment full of positivity and happi-
ness, which I am extremely grateful for. Additionally, my sisters Khushbu Kedia Lath and
Radhika Kedia have always been my strongest supporters and a constant source of encourage-
ment. I have been blessed with amazing relatives, who I know always keep me in their prayers.
In particular, I would like to thank Tulsian family and Goenka family, without whom I would not
be here, pursuing this education. Their blessings and constant support have played a very instru-
mental role in my career. I am grateful for the love and tolerance of Rohit Duhan. His kindness,
patience and sacrifices made it possible for me to complete my work without many hurdles. He
has been a key figure in this achievement.
This dissertation is dedicated to my mother Renu Kedia and my late grandmother Gita Devi Kedia, the two women who made all the difference in my life.
vii
TABLE OF CONTENTS
Placental ‘Omics’ Study to Understand the Pathogenesis of Preeclampsia .................................... i ABSTRACT ......................................................................................................................... ii ACKNOWLEDGMENTS ............................................................................................................. iv TABLE OF CONTENTS .............................................................................................................. vii LIST OF TABLES ........................................................................................................................ xi LIST OF FIGURES ...................................................................................................................... xii LIST OF ABBREVIATIONS ...................................................................................................... xiv 1. Introduction ......................................................................................................................... 1
1.1 Preeclampsia ............................................................................................................. 1 1.1.1 History ............................................................................................................... 1 1.1.2 Definition of PE ................................................................................................. 1 1.1.3 Epidemiology of PE........................................................................................... 2 1.1.4 Conditions believed to be involved in the pathogenesis of PE: ........................ 3 1.1.5 The role of the placenta in PE ........................................................................... 8
1.2 Biological mass spectrometry (MS) .......................................................................... 8 1.2.1 MS applications to complex biological specimens............................................ 8 1.2.2 Mass analyzers for biomolecules ..................................................................... 10 1.2.3 Quadrupole MS................................................................................................ 13 1.2.4 TOF .................................................................................................................. 15
1.3 Proteomics............................................................................................................... 19 1.3.1 Pre-separation before MS ................................................................................ 20
1.4 Lipidomics .............................................................................................................. 23 1.5 Tissue analysis ........................................................................................................ 26
1.5.1 Placental analysis ............................................................................................. 27 1.6 Importance of the LMW biomolecules and approaches to MS interrogation of
smaller biomolecules ................................................................................................................. 30 1.7 Layout of the dissertation........................................................................................ 35 1.8 References ............................................................................................................... 36
2. Novel ‘omics’ approach to study low abundance, low molecular weight components in a complex biological tissue: Regional differences between chorionic and basal plates of human placenta ....................................................................................................................... 44
2.1 Abstract ................................................................................................................... 44
viii
2.2 Introduction ............................................................................................................. 45 2.3 Materials and methods ............................................................................................ 49
2.3.1 Tissue collection .............................................................................................. 49 2.3.2 Tissue homogenization .................................................................................... 50 2.3.3 Protein depletion .............................................................................................. 50 2.3.4 Chromatographic separation ............................................................................ 51 2.3.5 Mass spectrometric analysis ............................................................................ 52 2.3.6 Time normalization and data analysis ............................................................. 52 2.3.7 Statistical analysis............................................................................................ 53 2.3.8 Normalization of candidate markers to reduce non-biological variation ........ 54 2.3.9 Identification of peptides of interest by tandem MS ....................................... 54 2.3.10 De novo sequencing ........................................................................................ 55 2.3.11 Identification of lipid markers......................................................................... 55
2.4 Results ..................................................................................................................... 56 2.4.1 Time markers to align chromatographic elution times .................................... 56 2.4.2 Differentially abundant low abundance, LMW weight biomolecules between
chorionic plate and basal plate. .............................................................................................. 56 2.4.3 Identification of peptides ................................................................................. 59 2.4.4 Identification of lipids ..................................................................................... 63
2.5 Discussion ............................................................................................................... 67 2.6 Conclusion .............................................................................................................. 73 2.7 References ............................................................................................................... 74
3. Global ‘Omics’ Evaluation of Human Placental Responses to Preeclamptic Conditions ........ 80 3.1 Abstract ................................................................................................................... 80 3.2 Introduction ............................................................................................................. 81 3.3 Materials and methods ............................................................................................ 82
3.3.1 Specimen collection ......................................................................................... 82 3.3.2 Sample processing ........................................................................................... 83 3.3.3 Explant culture ................................................................................................. 83 3.3.4 Homogenization of placental explants ............................................................ 84 3.3.5 Mass spectrometric analysis ............................................................................ 84 3.3.6 Statistical analysis............................................................................................ 85 3.3.7 Peptide identification ....................................................................................... 86 3.3.8 Lipid identification .......................................................................................... 87
3.4 Results ..................................................................................................................... 87
ix
3.4.1 Chromatographic time alignment .................................................................... 87 3.4.2 Low abundance, LMW weight species altered by PE conditions ................... 87 3.4.3 Sequenced peptides .......................................................................................... 95 3.4.4 Lipid classification .......................................................................................... 96 3.4.5 Condition specific signatures........................................................................... 97
3.5 Discussion ............................................................................................................. 100 3.6 References ............................................................................................................. 112
4. Targeted tissue ‘omics’ analysis comparing markers changed in culture in response to proposed mediators of preeclampsia to those markers in actual preeclamptic placentae .......... 119
4.1 Abstract ................................................................................................................. 119 4.2 Introduction ........................................................................................................... 120 4.3 Methods................................................................................................................. 121
4.3.1 Tissue collection ............................................................................................ 121 4.3.2 Specimen processing ..................................................................................... 123 4.3.3 Chromatographic separation .......................................................................... 123 4.3.4 Mass spectrometric analysis .......................................................................... 124 4.3.5 Data analysis .................................................................................................. 124 4.3.6 Statistical analysis.......................................................................................... 125
4.4 Results ................................................................................................................... 128 4.4.1 Chromatographic time alignment .................................................................. 128 4.4.2 Evaluation of 146 biomarkers that demonstrated differences under PE
conditions in PE placenta .................................................................................................... 128 4.5 Discussion ............................................................................................................. 135 4.6 Conclusion ............................................................................................................ 139 4.7 References ............................................................................................................. 140
5. Concluding remarks ................................................................................................................ 145 5.1 Summary of Current research ............................................................................... 145
5.1.1 Summary: Chapter 2 ...................................................................................... 145 5.1.2 Summary: Chapter 3 ...................................................................................... 146 5.1.3 Summary: Chapter 4 ...................................................................................... 146
5.2 Limitations of current research ............................................................................. 147 5.2.1 Limitations: Chapter 2 ................................................................................... 147 5.2.2 Limitations: Chapter 3 ................................................................................... 148 5.2.3 Limitations: Chapter 4 ................................................................................... 149
5.3 Future objectives ................................................................................................... 150
x
5.3.1 Future research objective: Chapter 3 ............................................................. 150 5.3.2 Future research objectives: Chapter 4 ........................................................... 151
5.4 References ............................................................................................................. 153
xi
LIST OF TABLES
Table 2.1. List of 16 differentially molecular species across all 10 time windows. 58
Table 2.2. Summary of the chemical identities of peptides differentially expressed in chorionic
plate and basal plate of human placenta. 61
Table 2.3. Summary of the chemical classes and components of the lipids differentially
expressed in chorionic plate versus basal plate of human placenta. 66
Table 3.1. List of differentially expressed biomolecules for all the three ‘stressed’ conditions. A.
Hypoxia, B. Hypoxia-reoxygentation, C. TNFα 90
Table 3.2. Condition specific changes in biomolecules 99
Table 3.3. Summary of categorized markers with their identifications and biological significance
102
Table 4.1. Mean and standard deviation of GA for preeclamptic and healthy pregnancies 122
Table 4.2. Biomarker with statistical significant differences between PE placenta and healthy
controls. 130
Table 4.3. Statistically significant markers with their identities and average elution time in the
total ion chromatogram 132
xii
LIST OF FIGURES
Figure 1.1. Abnormal placentation in PE. 7
Figure 1.2. Quadrupole time-of-flight mass spectrometer. 12
Figure 1.3. Schematic of Linear quadrupole. 14
Figure 1.4. Principle of operation for reflectron ion mirror (m1: m/z with higher velocity; m2: m/z
with lower velocity; Ekin:Kinetic energy). 18
Figure 1.5. Tandem mass spectrometry of peptides and fragment ion nomenclature. 22
Figure 1.6.Structures of few classes of lipids. R represent fatty acid chain A) Phosphocholines B)
Phosphoethanolamines C) Sphingomyelin D) Acylcarnitines 25
Figure 1.7. Molecular weight distribution of proteins in humans. 32
Figure 2.1. An overlay of 16 mass spectra in the region containing peptide m/z 718.36 with its
isotope envelope (z = +2). Red: tissue collected from chorionic plate; blue: tissue collected
from outer cotyledons (basal plate). This species was more abundant in the chorionic plate
of placenta (p=0.007). 57
Figure 2.2. Averaged MS2 spectra of a second differentially expressed peptide m/z 808.82 (z =
+2) (mass accuracy of the precursor peak =27 ppm) and the MS fragment-ion calculator’s
prediction of b and y ions. This peptide was identified as a fragment of phosphorylated
fibrinopeptide A. 62
Figure 2.3. Summed MS2 spectrum of a differentially expressed sodiated
lysophosphatidylcholine (m/z 544.33, z = +1) (mass accuracy of the precursor peak =51),
with its collision product ions and their proposed chemical structures or the proposed
component lost with fragmentation. Mass accuracies of individual product ions having m/z
xiii
values of 86.09, 104.1, 146.97, 184.06, 258.1, 339.26, 485.23 were 64 ppm, 59 pm, 68 ppm,
42 ppm, 37 ppm, 63 ppm, 53 ppm respectively. 65
Figure 2.4. Differential expression of peptide with m/z of 718.36 (z = +2). A. This is a line plot
of the abundance of the peptide in the basal versus the chorionic plate of each of the 12
replicates demonstrating the consistency of its differential expression. Each colored line
represents a single specimen. B. This is a box and whisker plot of the same peptide and its
mean abundance in the same two regions of the 12 placentas. 71
Figure 3.1. MS overlay and box-and-whisker plot for peptide m/z 827.78 (=+6) A. Overlay of 26
mass spectra in the region containing peptide m/z 827.78 (z=+6) with its isotope envelope.
Blue, placental explants under control conditions. Red, placental explants under hypoxic
conditions. Species less abundant in the hypoxic placental explants. B. Box-and-whisker
plot of peptide m/z 827.78 (z=+6) 89
Figure 3.2. Scatterplot of Group Membership Based on the Statistical Model. The numbers, 1, 2,
3, and 4, correspond to the Controls, Hypoxia, Inflammatory Cytokines, and Hypoxia-
Reoxygenation classes, respectively. The x, y, and z axes, which are not constrained to be
orthogonal, represent the predicted probabilities from the lasso model for the Controls,
Hypoxia, and Hypoxia-Reoxygenation treatment classes, respectively. 98
Figure 3.3. Venn diagram to represent the overlap in biomarkers between the three PE conditions
applied on placental explants. 111
Figure 4.0.1. MS/MS spectra of identified acylcarnitine (m/z 426.35; z=+1)) 134
xiv
LIST OF ABBREVIATIONS
µg/ml Micrograms per microliter µl Microliters 2-DGE Two-dimensional gel electrophoresis ACN Acetonitrile ACN Acetonitrile Ar Argon AT1-AA Angiotensin II type 1 receptor autoantibodies AUC Area under curve CHCA ɑ-cyano-4-hydroxycinnamic acid CI Chemical ionization CID Collision induced dissociation C-section Caesarian CT Cytotrophoblasts C-termini Carbon termini d.c. Direct current Da Daltons DHB 2,5-dihydroxybenzoic acid DMEM Dulbecco’s modified eagle medium E Kinetic energy ECD Electron capture dissociation EI Electron ionization ESI Electrospray ionization ETD Electron transport dissociation FbA Fibrinopeptide A FFPE Formalin-fixed paraffin embedded FTICR Fourier transform ion cyclotron resonance GA Gestational age GluFib Glu-1-Fibrinopeptide H/R Hypoxia-reoxygenation HCD High collision dissociation Hgb Hemoglobin HIF-2α Hypoxia inducible factor-2α HILIC Hydrophilic interaction liquid chromatography HMDB Human Metabolome Database HMW High molecular weight HPLC High performance liquid chromatography ICAT Isotope-coded affinity tag IEF Isoelectric focussing IR Infrared ITRAQ Isobaric tags for relative and absolute quantitation LASSO Least absolute shrinkage and selection operator LCM Laser capture microdissection LC-MS Liquid chromatography-mass spectrometry LLE Liquid-liquid extraction
xv
LMW Low molecular weight LOD Limit of detection LPC Lysophosphocholine m Mass m/z Mass-to-charge MALDI Matrix-assisted laser desorption/ionization MARS Multiple affinity removal systems MCP Microchannel plate MDA Malondialdehyde MS Mass spectrometry MS/MS Tandem mass spectrometry MudPIT Multidimensional protein identification technology N2 Nitrogen Na Sodium nM Nanomolars NP Non-polar N-termini Amino termini PBS Phosphate buffer saline PC Phosphatidylcholines PEt Phosphoethonolamines PE Preeclampsia pg Picograms PMT Photomultiplier tube PS Phosphoserine Psi Pounds per square inch PTM Post-translational modifications Q-TOF Quadrupole-time-of-flight r.f. Radiofrequency ROS Reactive oxygen species RP-LC Reversed phase liquid chromatography rpm Revolutions per minute SDS-PAGE Sodium dodecyl sulfate-polyacrylamide gel electrophoresis sFlt-1 Soluble fms-like tyrosine kinase-I SM Sphingomyelin ST Syncytiotrophoblasts STBM Syncytiotrophoblast microvesicles TIC Total ion chromatogram TMA Trimethylamine TNFα Tumor necrosis factor-alpha UV Ultraviolet v Velocity V Volts v/v Volume/volume VEGF Vascular endothelial growth factor XIC Extract ion chromatogram
1
1 Introduction
1.1 Preeclampsia
1.1.1 History
Headaches accompanied by heaviness and convulsions during pregnancy are considered
bad. This was the first suggestion of preeclampsia (PE) made by Hippocrates in 400 BC.1 In the
late 1800s the theory of causation was believed to be the involvement of toxins. Differentiation
of epileptic seizures from eclampsia, the extreme swelling in eclamptic women and the presence
of protein in urine were described by Bossier de Sauvages in 1739, Demanet in 1797 and Pierre
Rayer in 1840 respectively. After the invention of the mercury manometer in 1896, it was recog-
nized that PE was a hypertensive disorder. Although PE is still referred as “disease of theories”,2
recent advances have been made in understanding the pathophysiology of PE. This includes in-
volvement of the mother’s placenta as an essential and perhaps the central feature in the patho-
genesis of this disease.3
1.1.2 Definition of PE
The diagnosis of PE is based on the presence of de novo hypertension after 20 weeks ges-
tation and proteinuria. The current guidelines are as follows:
Blood pressure: An elevated systolic blood pressure > 140 mmHg and a diastolic pressure
of > 90 mmHg (>160 mmHg/110 mmHg is diagnostic of severe PE) on at least two readings, at
least 4 hr apart.4
2
Proteinuria: >300 mg protein/24 hr urine
If symptoms of PE remain untreated, the disease can progress to eclampsia5 defined as a
grand mal seizure in pregnant women, not related to any preexisting neurologic condition.
1.1.3 Epidemiology of PE
Occurrence of PE varies worldwide. The incidence ranges from 2-10+ % around the
globe.5 The World Health Organization’s estimate of PE incidence in developing countries is 7
times higher than the developed nations.5 PE is the second leading cause of maternal deaths ac-
counting for over 75,000 mothers6 and claiming lives of over 500,000 infants.7 As a consequence
of PE, up to 12% of all newborns are severely growth restricted and one-fifth of them are born
preterm.8 Symptoms of PE resolve quickly after delivery of the baby in almost all cases.8 Long
term sequelae of PE remains under investigation, but include a higher risk of hypertension,
stroke and ischemic heart disease for mothers.8 Children born preterm or small in size have a
higher risk of cerebral palsy.8 Several risk factors are proposed in case of PE. For primigravidas,
a family history of PE increases the risk of a severe PE by 4 fold.9 In addition, it is also believed
that a man who has fathered one preeclamptic pregnancy is more likely to father another preg-
nancy complicated by PE as well, even with a separate partner.10 In comparison to multiparous
women, nulliparous women are up to 3 times more likely to develop PE.10 Several preexisting
conditions like hypertension, diabetes mellitus can increase the risk of this disease.10 Smoking is
associated with a reduced risk of PE, although the adverse effects of smoking still outweigh the
potential benefits during a pregnancy.10
The pathophysiology of PE remains poorly understood. It has been widely accepted that
the onset of the disease process starts with impaired spiral artery remodeling, as a consequence
3
of inadequate trophoblast invasion.11 Trophoblasts represent the outermost layer of the blasto-
cyst. These cells form a major portion of the placenta providing nutrients to the embryo. In hu-
mans, on around the 8-9th day after conception the blastocyst with its trophoblast cells begins to
invade the endometrium (Figure 1.1). This step is crucial for a healthy pregnancy.12 Trophoblast
cells are differentiated into undifferentiated cytotrophoblasts (CT) and highly differentiated syn-
cytiotrophoblasts (ST).13 ST are the cells in direct contact with the maternal blood and invade the
endometrium (interstitial invasion) and its blood vessels (endovascular invasion). These cells
thus play an important role in the remodeling of spiral arteries. Spiral arteries are responsible for
providing blood supply to the endometrium during pregnancy.14 These arteries, as a consequence
of the invasion are converted from high-resistance blood vessels to low-resistance high flow ar-
teries. This conversion ensures a sufficient amount of blood flow to meet the high metabolic de-
mand of the fetal-placental unit during pregnancy. It is believed that PE starts as a consequence
of an incomplete remodeling of the maternal spiral arteries causing reduced blood flow to the
placenta (Figure 1.2).15 This reduced blood flow has consequences, such as a hypoxic placenta
and/or an increased state of placental oxidative stress.
1.1.4 Conditions believed to be involved in the pathogenesis of PE:
Many abnormalities accompany established, clinically evident PE. However, far fewer
features of the disease have been found to precede the maternal features of PE. Among these, a
few have been documented in several studies making them better established and more accepted
as early and likely contributory changes.16 Among these abnormalities, those that are now con-
sidered to be etiologic or highly involved in the disease process are as follows:
4
1.1.4.1 Hypoxia/ischemia of placenta
Hypoxia, in particular a hypoxic placenta is likely the most commonly proposed mediator
of PE. Due to impaired spiral artery remodeling, blood flow to the placenta is reduced, leading to
the placenta being deprived of oxygen (Figure 1.3). The hypoxic placenta is then believed to re-
lease vasoactive factors into the maternal circulation causing endothelial dysfunction and ulti-
mately the symptoms of PE.17 Proposed mediators include soluble fms-like tyrosine kinase-1
(sFlt-1), pro-inflammatory cytokines, and angiotensin II type 1 receptor autoantibodies (AT1-
AA).17 In addition, studies18 have been conducted that show a higher expression of hypoxia in-
ducible factor-2α (HIF-2α) in preeclamptic placenta, which is a paracrine factor that regulates
several pathways to respond to the oxygen environmrent to preserve cellular functions.18
1.1.4.2 Oxidative stress
Oxidative stress can occur as a consequence of an imbalance between production of reac-
tive oxygen species (ROS) and antioxidant defences. Oxidative stress of the placenta is believed
to occur as a result of an intermittent supply of blood to placenta due to impaired spiral artery re-
modelling.19 This hypoxia-reoxygenation causes placental oxidative stress, indicated by an in-
creased concentration of malondialdehyde (MDA) in preeclamptic placenta, a product of ROS
activity and a marker of lipid peroxidation.20 A higher concentration of placental tumor necrosis
factor-α (TNF-α), a pro-inflammatory cytokine, has also been reported in response to increased
ROS.20 Increased secretion of this cytokine into the maternal circulation could lead to peroxida-
tion of maternal lipids and in turn cause endothelial dysfunction.20 Oxidative stress of placenta is
also accompanied by a decreased expression of superoxide dismutase mRNA and protein expres-
sion potentially limiting its oxidant scavenging activity.21 This reduction may also be linked to
5
an increase in apoptotic activity which can cause microvillous fragments from placenta to be re-
leased into the maternal circulation further leading to endothelial dysfunction and presumably
symptoms of PE.22
1.1.4.3 Angiogenic factors
There is substantial evidence for a dysregulation of pro-angiogenic and anti-angiogenic
factors in many women with established PE and a suggestion of changes in circulating levels of
these factors up to 5 weeks prior to PE in some women destined to develop the disease.23
Placental growth factor (PlGF) and vascular endothelial growth factor (VEGF) are pro-an-
giogenic factors essential for placental vascular development. It has been proposed that PE oc-
curs due to an imbalance of these fundamental angiogenic factors.15 Studies have demonstrated a
decrease in free levels of PlGF and VEGF15 and an increase in sFlt-1 in the circulation of women
with PE24 as compared to controls. Reduced amounts of pro-angiogenic factors may result in im-
paired spiral artery remodeling and thus lead to features of PE. A study conducted by Wang et al
demonstrated a decreased invasiveness by CT in response to an increase in sFlt-1 in vitro.25 sFlt-
1 which is a splice variant of the VEGF receptor Flt-1, blocks the actions of both VEGF and
PlGF by binding them in the circulation. This binding prevents the interaction of these mediators
with their receptors, thus attenuating their angiogenic properties.15
1.1.4.4 Cytokines
Among other factors thought to contribute to the pathogenesis of PE are increased levels
of certain pro-inflammatory cytokines. Multiple studies have shown an increase in circulating
6
levels of this class of molecules in preeclamptic patients when compared to normotensive moth-
ers.26
Other studies have shown an increase in the mRNA and protein expression of inflamma-
tory cytokines in preeclamptic placentas.27 This upregulation of cytokine production is thought to
be caused by the focal regions of hypoxia that result from shallow vascular invasion during im-
plantation. The hypoxic environment is sensed by HIF-alpha, which directly regulates the tran-
scription of certain cytokines such as tumor necrosis factor α (TNFα). This finding has been con-
firmed by other studies in which elevated amounts of TNFα were generated when placental ex-
plants were placed in hypoxic conditions.28
Although these and many other abnormalities have been proposed as causes of PE, mech-
anistic involvement of any of these conditions has yet to be proved . Nevertheless, placental in-
volvement is almost certain in the pathogenesis of PE.
7
Figure 1.1. Abnormal placentation in PE. (Reprinted with permission from Wang, A.; Rana, S.; Karumanchi, S. A., Preeclampsia: the role of angiogenic factors in its pathogenesis. Physi-ology (Bethesda) 2009, 24, 147-58.)
8
1.1.5 The role of the placenta in PE
The placenta is the temporary organ that performs crucial functions in sustaining a suc-
cessful pregnancy. It is responsible for the import of nutrients, exchange of gases and removal of
waste products from the fetus. In addition, it synthesizes several peptides and synthetic hormones
regulating many functions required for a healthy pregnancy. Symptoms of PE resolve rapidly af-
ter delivery which could either be due to the involvement of placenta or the fetus.19 Placenta has
been widely implicated in the pathogenesis of PE, but not the fetus. Moreover, PE has also been
observed with a molar pregnancy.19 A molar pregnancy is a condition, where the trophoblastic
tissue supposed to become fetus develops into an abnormal growth or mass. Occurrence of PE in
such circumstances indicates the involvement of the placenta and placental factors and not the
fetus in the pathophysiology of the disease.19
1.2 Biological mass spectrometry (MS)
1.2.1 MS applications to complex biological specimens
A MS is a device that measures the mass-to-charge (m/z) ratio of gas-phase ions. Appli-
cation of MS to biological samples increased exponentially with the development of “soft ioniza-
tion sources” such as Electrospray Iionization (ESI) and Matrix-Assisted Laser Desorption/Ioni-
zation (MALDI). Both of these ion sources convert analytes into gas-phase ions with minimum
in-source fragmentation and produce [M + zH]z+ ions. This is in contrast to more conventionally
used ionization techniques such as Chemical Ionization (CI) and Electron Ionization or Electron
Impact (EI) which cause a higher degree of in-source fragmentation.29
9
1.2.1.1 ESI
ESI for the analysis of biological macromolecules was introduced by Fenn and co-work-
ers in 1984. In this technique, a volatile liquid containing the biomolecules of interest as well as
containing a weak volatile acid is introduced at a flow rate ranging from one to several hundred
µl/min through a very narrow capillary (~0.2 mm o.d and ~0.1 mm i.d) made of stainless steel or
quartz silica. This capillary is maintained at a high voltage (2-6 kV) which causes aerosol for-
mation in the form of a Taylor cone with production of charged droplets.30 Each charged droplet
initially is in the micrometer range and positively charged due to the presence of H+ and other
cations such as Na+ and K+. Droplet formation is followed by solvent evaporation by a coaxial
flow of heated sheath gas (e.g. dry nitrogen gas). This heating causes evaporation of the solvent
from the charged droplets. As a consequence, the charge density keeps increasing and ap-
proaches the Rayleigh limit where the surface tension of the liquid droplet is balanced by the
coulombic repulsion.29 Increasing solvent evaporation and consequent repulsive forces causes a
coulombic explosion and release of progeny droplets.31 This process continues until gas phase
ions are formed from these droplets. These ions are then directed to the mass analyzer towards an
oppositely charged plate.
1.2.1.2 MALDI
In MALDI the analytes are co-crystallized with a suitable matrix molecule. Examples of
organic matrices include 2,5-dihydroxybenzoic acid (DHB) and ɑ-cyano-4-hydroxycinnamic
acid (CHCA). These matrices have a very strong absorption in the ultraviolet (UV) or infrared
(IR) range but remote from protein UV absorbances. The laser (e.g. nitrogen laser) is fired on the
matrix which absorbs the energy of the laser light and heats the matrix molecules very rapidly,
10
causing the specimen to be simultaneously desorbed and ionized. The mechanism of ionization
for MALDI is still debatable, but it is believed32 that the desorbed matrix molecules transfer a
proton to the analyte thus charging it, typically with one positive charge per analyte molecule.
These charged analytes are then accelerated into the vacuum region of the mass analyzer and di-
rected into the MS.32
A potential advantage of ESI over MALDI, which more commonly produces [M+H]1+
species, is the ability of ESI to produce multiply charged species. This is especially important in
the field of proteomics as large proteins can be multiply charged and can fall in the mass range of
the mass analyzers such as quadrupole-time-of-flight (Q-TOF), commonly used to study proteins
and peptides.
1.2.2 Mass analyzers for biomolecules
Mass spectrometers (MS) commonly used for the analysis of biomolecules are Quadru-
pole MS, Time-of-flight MS, Quadrupole ion trap MS, Orbitrap MS, and Fourier transform ion
cyclotron resonance (FTICR) MS.
For my research, the mass analyzer used was a tandem MS, a hybrid of two mass analyz-
ers: a quadrupole coupled to a time-of-flight MS (Q-TOF).
In a QqTOF-MS instrument there are several sectors including a mass-resolving quadru-
pole (Q), a radiofrequency (r.f.) quadrupole (q), and the mass analyzer (TOF). Most of the com-
mercially available QqTOFs have an additional Q0 which can be a hexapole or octopole and is
operated in a r.f. only mode to act as an ion focusing guide.33 So for QqTOF-MS, the configura-
tion is comparable to a triple quadrupole MS configuration, except the final quadrupole is re-
placed by a TOF detector (Figure 1.4). Replacement by a TOF mass analyzer provides certain
11
advantages such as 1) higher resolution, achieved by to ion mirrors and a delayed extraction pro-
cess, 2) higher mass accuracy, 3) and parallel recording of all ions.
12
Figure 1.2. Quadrupole time-of-flight mass spectrometer. (Reprinted with permission from Chernushevich, Igor V., Alexander V. Loboda, and Bruce A. Thomsoken from. T "An introduction to quadrupole–time‐of‐flight mass spectrometry." Journal of Mass Spectrometry 36.8 (2001): 849-865.)
13
1.2.3 Quadrupole MS
A quadrupole MS consists of four parallel hyperbolically or cylindrically shaped metal
rods at equal distances from each other. These electrodes extend in the z direction and are ar-
ranged in a square configuration. Each set of diagonal rods carry identical charge while the other
set carries the opposite charge (Figure 1.5). As an ion enters in z direction, it gets attracted by the
rod with an opposite potential, since an r.f. (radio frequency) voltage is applied between diagonal
rods, the sign of the potential changes and thus the ion forces attraction and repulsion in x and y
direction. For the quadrupole mass filter a d.c. (direct current) voltage is superimposed on the ap-
plied r.f voltage. For a given ratio of r.f and d.c voltages, only a certain m/z or a certain range of
m/z’s will have stable trajectory between the quadrupole region to pass through it.33
14
Figure 1.3. Schematic of Linear quadrupole. (Reprinted with permission from Gross, J. Mass Spectrometry, 2nd ed.; Springer, 2011; pp 147)
15
The ‘q’ in QqTOF act as a collision cell to cause fragmentation. This technique is re-
ferred as tandem mass spectrometry (MS/MS) (Figure 1.6). The collision cell works in the r.f.
only mode to transmit all m/z values with stable trajectories. In MS/MS mode, Q1 works as a
mass filter and transmits only a desired parent or precursor ion with the pre-determined m/z
value. This parent or precursor enters the collision cell or the ‘q’ sector where it undergoes colli-
sion-induced dissociation (CID) or breakage. CID is a process where the precursor or parent of
interest is accelerated with high energy (generally between 15-200 eV) and collides with a neu-
tral gas, e.g. nitrogen or argon. As a consequence of this collision some of the kinetic energy of
the molecule is converted into internal energy and causes bond breaking and fragmentation of the
targeted ion. For proteins and peptides the site of cleavage is known for CID and is discussed in
more detail later in this dissertation. The first set of fragments produced in MS/MS is also known
as the MS2 spectrum and one can perform additional fragmentation of a selected MS2 product to
generate MS3 products and so on using advanced instruments such as Orbitrap. During MS1 con-
ditions where no MS/MS is desired, parameters are similar, only the collision energy is kept be-
low 10 eV to prevent fragmentation.33
1.2.4 TOF
The TOF MS acts as mass analyzer where the m/z of the analyte ion is determined by the
flight time in a field-free drift tube of a determined length. As the ions leave the collision cell,
they are accelerated again, in a parallel beam focused by the ion optics and reach the ion modula-
tor/ion pulser of the TOF. At this point a high voltage pulse is applied which forces the ions to
move perpendicularly and enter the field-free flight tube. At the end of this flight tube is an elec-
16
trostatic ion mirror which improves mass resolution and sends back ions from the ion pulser to-
ward the detector. The detector is made up of a microchannel plate (MCP), scintillator and a pho-
tomultiplier tube (PMT). The first stage of the MCP is a thin plate made up of multiple micro-
scopic channels with approximately 10 micrometer diameters. As each analyte ion hits the MCP
plate, a pulse of approximately 10 electrons are produced which then hit the scintillator resulting
in photon emission. These photons are focused to the PMT which amplifies them producing an
electrical signal proportional to the number of photons.33
Time-of-flight for each ion is dependent on its mass. As a result, each ion has a unique
flight time. Velocity of a given ion is determined by using the following formula: v = √(2E/m),
where E stands for a given kinetic energy, m for mass and v for velocity. Thus, according to the
equation lighter ions will have a higher velocity and heavier ions, a lower velocity. This causes
their separation in the field-free flight tube.33
1.2.4.1 Approaches developed to increase the resolution of TOF instruments
Delayed extraction: During certain ionization methods like MALDI, there is a significant energy spread of the
produced ions causing a significant temporal distribution and thus loss of resolution. To address
this issue, a process known as delayed extraction is employed where a delay occurs between gen-
eration and acceleration of ions. Due to the pulsed nature of MALDI, ions with the same masses
move with different velocities. As can be seen in Figure 1.7, the ion moving with higher velocity
(red) experiences a lesser extraction voltage between target and extraction plate (+V) as com-
pared to the ion moving slower (blue). The slowest ion thus receives maximum energy therefore
compensating for temporal distribution and improving the resolution.3
17
Reflectron/ion mirror A reflectron consists of decreasing electric field and is located behind the field free re-
gion of the TOF tube. These mirrors are stacks of electrodes with increasing voltages, responsi-
ble for compensating for the difference in velocities, i.e. kinetic energies, and thus improving
resolution. Ions with higher velocities penetrate deeper into the mirror assembly until their K.E
becomes zero. Ions with more K.E move deeper and thus spend more time in the reflectron than
an ion of the same mass but with low velocity, this corrects for the difference in velocities
thereby increasing the resolution (Figure 1.8).35
18
Figure 1.4. Principle of operation for reflectron ion mirror (m1: m/z with higher velocity; m2: m/z with lower velocity; Ekin:Kinetic energy). (Reprinted with permission from Gross, J. Mass Spectrometry, 2nd ed.; Springer: Germany, 2011; pp 126)
19
1.3 Proteomics
The term “proteomics” was first coined in 1995 and was defined as “the large-scale char-
acterization of the entire protein complement of a cell line, tissue, or organism”.36 The emerging
field of proteomics has a wide range of applications ranging from discovering biomarkers for
various diseases to finding new targets for drug therapy. The phenotype of a cell is more closely
represented by its proteome as compared to its genome. While transcriptional studies yield useful
information about gene expression patterns in cells, proteomic studies may provide important ad-
vantages.37 Proteins themselves are the direct participants in the cellular processes of both
healthy and diseased tissues; indeed, recent studies have indicated that mRNA levels have only
moderate power to predict protein concentrations. In fact, it is estimated that only about 40% of
protein expression levels can be explained by mRNA expression levels.37 Hence, proteomic
methods, which more directly measure protein abundance, likely provide a clearer picture of the
pathophysiology of disease conditions such as PE. In addition, most transcriptomic technologies
are incapable of predicting splice variants arising from the same gene, and none can predict post-
transcriptional modifications that may affect the function of a protein.38 These limitations pre-
vent mRNA studies from investigating important changes involved in various diseases and argue
for more extensive proteomic studies in cells and tissue.
Proteomics has been a challenging field considering the dynamic range of protein con-
centrations in an organism. For example, dynamic range of protein concentration in human se-
rum is approximately 10 orders of magnitude.39 Current technologies available cannot address
this issue and analyze the complete proteome of a cell or a tissue.
20
1.3.1 Pre-separation before MS
Due to the complexity of biological samples and the wide dynamic range of protein con-
centrations, a pre-separation step is usually employed before introducing the samples into the
mass spectrometer. The two most commonly used approaches to achieve pre-separation are 1)
Two-dimensional gel electrophoresis (2-DGE) and 2) Reversed phase liquid chromatography
(RP-LC).40 2-DGE uses isoelectric focusing (IEF) in the first dimension which separates the pro-
teins based on their isoelectric point and by molecular weight in second dimension using sodium
dodecyl polyacrylamide gel electrophoresis (SDS-PAGE). Each spot on the gel represents one
protein which can then be cut out of the gel and digested using enzymes such as trypsin. This di-
gestion yields a defined set of smaller peptides from the parent protein which are then analyzed
using a suitable mass spectrometer.40 The second approach uses a reversed phase/hydrophobic
column to separate protein mixtures based on the differences in the hydrophobicities of individ-
ual proteins. This is currently a more commonly used approach as it has several advantages over
2-DGE such as better separation of a broad range of molecules, better sensitivity, easier opera-
tion and more compatible hyphenation with the mass spectrometers. Reversed phase columns
commonly used are: C-18 (18-Carbon chain), C-8 (8-Carbon chain) and C-4 (4-Carbon chain).
The longer the carbon chain, the more hydrophobic is the column.
Structural elucidation of proteins is extremely important in order to gain more infor-
mation on that species. Protein sequencing is the process of determining the order of amino acids
beginning from the N-termini (amino termini) to the C-termini (carboxy termini). In earlier days,
sequencing was achieved by performing Edman degradation to identify N-terminal amino acid
sequences. With the recent advances in MS-based techniques and increasing computing power,
sequencing by Edman degradation has been replaced by MS-MS sequencing.
21
The QqTOF-MS most often uses CID as the ion activation/fragmentation method to se-
quence and identify proteins and peptides of interest. As described earlier, collisions break the
most labile bond which in case of proteins and peptides is the amide bond, (-CO-NH-). As a con-
sequence, CID yields b- and y-ion series which is represented in Figure 1.9. The b-ions represent
the amino terminal of the protein, while the y-ion series represent the peptides starting at the car-
boxy terminus. Different collision energies can be required to provide a more complete and in-
formative MS/MS spectra. Spectra obtained from all these collision energies can then be aver-
aged using the MS system’s software and this averaged spectra can then be submitted to search
engines like MASCOT, SEQUEST, Spectrum Mill to finally determine the identity of the target
parent protein or peptide.41
There are instances when none of the protein search engines will provide a match for the
identification. In such situations, a technique known as de novo sequencing is applied. This is a
process of determining the order of amino acids in the protein without the help of a database. De
novo sequencing involves determining the N-termini or the C-termini and moving from there to
establish the amino acid sequence based on the differences in the masses between b-series and/or
y-series peaks. This is a challenging and a time consuming approach, especially for high charged
state, post-translationally modified proteins/peptides.42
22
Figure 1.5. Tandem mass spectrometry of peptides and fragment ion nomenclature (Reprinted with permission from Zhurov, K. O.; Fornelli, L.; Wodrich, M. D.; Las-kay, U. A.; Tsybin, Y. O., Principles of electron capture and transfer dissociation mass spectrometry applied to peptide and protein structure analysis. Chemical Society reviews 2013, 42 (12), 5014-30.
23
1.4 Lipidomics
Lipidomics is a relatively new field of study and was first reported by Han and Gross in
2003. It has been defined as “the full characterization of lipid molecular species and of their bio-
logical roles with respect to expression of proteins involved in lipid metabolism and function, in-
cluding gene regulation”.43 Lipids have important cellular functions, mostly energy storage,
structural stability and cellular signaling.44
Lipid analysis, similar to proteomics can be broken down into following components:
sample extraction, MS data acquisition and data analysis.
Since lipids form a diverse group of molecules with varying polarities and structures, a
sample extraction approach which can extract this wide range is desirable. The most common
way of extracting lipids has been by liquid-liquid extraction (LLE) using the Bligh and Dyer
method. This approach uses a 1:2 v/v chloroform:methanol mixture which allows the more non-
polar lipids to be extracted into the chloroform and the relatively polar ones into the methanol
present.45 Another approach to study lipids is by performing organic solvent precipitation with
acetonitrile (ACN) or methanol. The limitation of this approach is that it extracts only a small
range of lipid classes. For example this would include some glycerophospholipids like phospha-
tidylcholines (PC) and phosphoethanolamines (PEt) (Figure 1.10).45
Data acquisition of extracted lipids can be performed by either: 1) Direct injection into
the MS 2) Liquid chromatography-mass spectrometry (LC-MS) analysis.
Direct injection is a simple technique which allows for faster analysis and detection of
several lipid classes in a single MS run. This technique has some shortcomings such as determi-
nation of isobaric species, significant numbers of overlapping peaks and ion-suppression from
24
the presence of readily ionizable lipids. LC-MS on the other hand minimizes some of these ef-
fects. Columns commonly used for studying lipids by LC-MS are reversed phase columns sepa-
rating lipids based on their hydrophobicities, which mostly depends on the number of carbons.
Other less commonly used columns are normal phase columns (NP) and hydrophilic interaction
LC (HILIC) both separating lipids based on their hydrophilic functional groups, mainly residing
in their headgroups.45 For ionization of lipid compounds through ESI, both the positive and nega-
tive ion mode can be used to target different classes of lipids for MS analysis. Structural elucida-
tion of lipid classes can be challenging owing to the lack of knowledge and information regard-
ing their fragmentation mechanisms and limited in-silico MS/MS databases. The process of lipid
identification is very similar to de novo sequencing of proteins and peptides. CID is the most
common way to break a lipid molecule in the collision cell. Certain lipid classes have predicta-
ble MS/MS spectra and thus can be identified based on signature peaks or specific neutral losses.
For example PC’s MS/MS spectrum shows a peak at m/z 184.06 produced due to the headgroup
phosphocholine while PE have an abundant peak at the neutral loss of 141.01 due to loss of the
phosphoethanolamine headgroup. There are several databases which can help in the identifica-
tion of lipid classes, but no single database is complete. LIPID MAPS, METLIN, HMDB (Hu-
man Metabolome Database) are some of the commonly used databases for such predictions.
25
Figure 1.6.Structures of few classes of lipids. R represent fatty acid chain A) Phos-phocholines B) Phosphoethanolamines C) Sphingomyelin D) Acylcarnitines
26
1.5 Tissue analysis
Tissue analysis has several inherited advantages over other biological matrices. Several
orders of magnitude higher concentrations of tissue specific biomolecules and a closer represen-
tation of molecules in their native surroundings are some of these advantages.46 Challenges en-
countered while performing tissue analysis are tissue heterogeneity (overlap of different cellular
types),47 limited tissue availability in fresh form or a small amount of tissue not sufficient to per-
form the study. Some tissues are difficult to access and may limit studies to animals and may re-
quire sacrifice of the animal to obtain tissue preventing serial sampling. The most common ap-
proach for tissue analysis is tissue homogenization which involves grinding or cutting a part or a
whole tissue using different techniques like ball mill homogenizer, a mortar and pestle, a
blender, etc. Homogenization does not solve the problem of cellular heterogeneity. To address
this, researchers use laser capture microdissection (LCM) to target a specific cell population.47
Imaging techniques such as MALDI imaging have been attempted to study the protein repertoire
of intact tissues. These techniques are well-suited for studies requiring analysis of biomolecules
in their native architecture.48 MS remains the most useful technique to study biomolecules from
tissue samples processed using some of the described approaches. Preparation of formalin-fixed
paraffin embedded (FFPE) tissue is the traditional way to preserve pathological and clinical
specimens for long periods of time at room temperature. This technique crosslinks proteins thus
rendering their analysis very challenging using current analytical platforms especially MS,49 alt-
hough with the help of recent developments, it has become feasible to analyze proteins from
FFPE tissues using MS based techniques.49 However, it is recognized that such studies can only
be qualitative and will typically be limited in the range of proteins accessed.
27
Preeclampsia (PE) is proposed to be a disease of placenta. Hence, I will focus more on
placenta as a tissue for analysis of biomolecules using MS based tools.
1.5.1 Placental analysis
1.5.1.1 Placental proteomics
Placenta is a vital organ that mediates selective transfer of solutes and gases from the
mother to her growing fetus. It is known to perform a vast number of functions for maintaining a
healthy pregnancy, and this is supported by the fact that more than 20,000 DNA sequences are
expressed by a human placenta.50 Thus, exploring the protein expression patterns of such a vital
organ may give deep insights into the functions of placenta and shed light on placental diseases
such as PE.
Placenta is a highly complex organ exhibiting significant anatomical variation.50 Placen-
tal proteomics is in its infancy owing to the complexity of the organ and lack of available protein
isolation protocols. Several studies have been performed to investigate the proteome and sub-
proteome of placenta and its cells. J.M Robinson et al. in their review on placental proteomics
mention two approaches that can provide useful information about the placental proteome.51 The
first one referred to as ‘parallel profiling’ is a technique in which a common fractionation step is
used to isolate proteins and look for differences in the protein expression patterns between the
comparison groups (e.g. cases vs. controls). This approach has provided the most useful infor-
mation in terms of differentially expressed proteins in any diseased state when compared to nor-
28
mal condition. The second one focuses more on a defined sub-proteome of placenta. Here, pro-
tein profile of an enriched cellular fraction of the organ is established to dig deeper into functions
of placenta. I will now briefly discuss these approaches.
Parallel profiling
Parallel profiling is a comparison based approach which has the advantage of providing a
broader coverage of proteomic information. A substantial amount of research has been con-
ducted to compare the placental proteome of a normal placenta to a diseased one using the con-
cept of parallel profiling. Examples include a study conducted by Magnus and Stefan et al. to ex-
amine differences in proteins between preeclamptic and normal placentae.52 They found an accu-
mulation of apolipoprotein A1 in preeclamptic placenta, indicating a role of this protein in patho-
genesis of this disease. In another study with placenta as the target tissue, it was shown that an
increased amount of dynactin p-50 may be involved in the pathophysiology of PE. This was con-
firmed by the presence of higher amounts of anti-dynactin p-50 in the serum of preeclamptic
mothers.53 Both these studies as well as several other studies of the placental proteome have used
2-DGE as the method to separate proteins before trypsin digestion and MALDI MS. A more
‘global’ proteomics study was performed by Wang et al. in which a homogenate of placental tis-
sues was submitted to LC/MS/MS and 171 proteins were found to be differentially expressed be-
tween normal healthy placenta and preeclamptic placenta. 147 proteins were found to be down-
regulated while 24 proteins showed a higher expression in PE placentas. Examples of these pro-
teins included endoglin, ceruloplasmin, mitochondrial superoxide dismutase and transforming
growth factor beta.54 All the above mentioned studies were limited to abundant proteins using
bottom-up approaches.
29
Sub-proteome studies
As mentioned above, a second approach to carry out a proteomics study on placenta is to
isolate a sub-cellular fraction from the organ and extract proteins. Due to the ‘dynamic range
problem’ identification of the entire placental proteome becomes a challenging task. Anatomical
variations in placenta make this problem even worse. An alternative approach to reduce this
problem is to identify proteins in a defined sub-cellular fraction. This approach also increases the
probability of identifying low copy numbers, which fall below the detection limit of conventional
techniques.
Before getting deeper into the sub-proteome studies done to date, it becomes important to
know about the highly specialized cellular structures of human placenta, including the CT and
ST. The trophoblast cells, upon contact with uterine walls differentiate into CT. These cells then
differentiate into a highly specialized network of cells called ST, in which several cells fuse to
form a single cell or a multi-nucleated syncytium housed within a single plasma membrane.55
This epithelial structure forms the interface between maternal blood and placenta, and is consid-
ered to be the most important element in the selective transfer of substances across placenta to
the fetus. Due to its importance in pregnancy, many attempts have been made to enrich this por-
tion of placenta and explore its protein repertoire. A study conducted by Epiney et al. studied
proteins secreted from cytotrophoblast cells from both normal and preeclamptic placenta. A la-
bel-free approach was used to find 21 proteins with differential expression between the compari-
son groups.56 In another study, a relatively new approach LCM was used to isolate trophoblastic
cells from preeclamptic and normal placental samples. This group found that 31 of the 70 altered
30
proteins were downregulated in preeclamptic placenta while the remaining 39 were upregu-
lated.57 Both of these techniques used enzymatic digestion to perform bottom-up proteomics us-
ing LC/MS platforms.
1.5.1.2 Placental lipidomics
Lipidomics is a relatively new technique and placental lipidomics even newer. Only a
handful of studies have been published to study the altered lipidome in preeclampsia using pla-
centa as the study model. Henri et al. found an altered lipidome in preeclamptic placenta as com-
pared to controls.58 Placenta obtained from both the groups were homogenized and resuspended
in water, lipids from this sample were extracted using the Bligh-Dyer protocol and were ana-
lyzed by MALDI/TOF-TOF. Phosphoserine (PS) followed by macrolides/polyketides were the
most prevalent class of molecules with increased abundance in preeclamptic placenta when com-
pared to controls. Another lipidomics study performed by Baig et al. investigated the lipidome of
placental syncytiotrophoblast microvesicles (STBM) of preeclamptic placenta. This approach
also used the Bligh-Dyer methodology followed by LC/MS. Approximately 200 lipids were
quantified in STBM. According to this study, the major class of lipids observed in STBM were
sphingomyelins (SM) which were upregulated in the preeclampsia group.59
1.6 Importance of the LMW biomolecules and approaches to MS interrogation of smaller bio-
molecules
Most of the proteomics studies discussed above were performed using two 2-DGE to ana-
lyze proteins initially. Although it represents the most commonly used technique, 2-DGE is bi-
31
ased towards the identification of highly abundant, higher molecular weight proteins. LMW pro-
teins are important participants in several biological processes such as ribosome formation, cellu-
lar signaling and stress adaptation. This fraction of the proteome has been identified as having
important regulatory compounds such as cytokines, peptide hormones and chemokines. It is be-
lieved that 50% of the human proteome is composed of proteins below 26.5 kDa (Figure1.11).64
32
Figure 1.7. Molecular weight distribution of proteins in humans. (Reprinted with per-mission from Bäckvall, H.; Lehtiö, J. The Low Molecular Weight Proteome: Methods and Protocols,; Springer: New York, 2013)
33
LMW proteins despite their importance in the pathophysiology of various biological pro-
cesses are underrepresented in proteomic studies owing to the very few approaches designed to
isolate and focus on them. 2-DGE has a low sensitivity and usefulness for this fraction of the
proteome as these small proteins have a tendency to stain poorly and migrate out of the gel eas-
ily.64 Removal of abundant proteins makes the analysis of LMW species more feasible, espe-
cially if one is using MS based techniques. Some obvious advantages of protein depletion are
narrowing of the wide dynamic range of human proteome and minimizing problems associated
with ion suppression.
Several techniques have been used to deplete abundant, high MW proteins from biologi-
cal matrices.
One of the techniques uses an immunodepletion affinity column to deplete the most
abundant proteins from samples e.g. plasma. Some of these columns are named ‘Multiple affinity
removal systems’ (MARS) which remove the 7 most abundant proteins, the ‘MARS-7’ column
which removes the 14 most abundant proteins, the ProteoPrep20 column which removes the 20
most abundant proteins.60 These columns use a mixture of antibodies against the most abundant
proteins in plasma/serum samples to capture and retain those proteins thus reducing sample com-
plexity and eliminating the most highly abundant species reducing ion suppression.61 One limita-
tion of using an affinity-based column is that the many LMW proteins/peptides carried by big
proteins such as albumin cannot be studied as they are lost with the carrier proteins in this pro-
cess as well.
Another approach to reduce the dynamic range problem and deplete abundant proteins is
by performing centrifugal ultrafiltration. This technique can involve using different solvents to
release the bound small proteins/peptides from carrier protein,62 but other studies have shown
34
that approach using acetonitrile (ACN) as the organic solvent provides a more extensive in-depth
analysis of the LMW proteome from serum samples.63
Organic solvent precipitation has been used traditionally to precipitate abundant large
proteins. ACN is the most common solvent employed, which has been shown to precipitate most
of the HMW proteins while releasing the LMW peptides and other species from carrier proteins
including some classes of lipids. As a consequence the dynamic range becomes smaller and
many more LMW species are available to investigate because ion-suppression has been mini-
mized in MS analysis.
35
1.7 Layout of the dissertation
This dissertation is divided into 5 chapters. Chapters (2-4) involve working with human
placenta, starting from development and optimization of a tissue proteomic approach to studying
placentae cultured under preeclamptic conditions as well as the ones collected from actual
preeclamptic pregnancies. Chapter 5 concludes the dissertation with a brief discussion on poten-
tial pitfalls as well as future prospects of this study.
Chapter 2 describes an initial study conducted to optimize a tissue proteomics approach to
investigate differences in biomolecules between two regions of human placenta collected from
healthy pregnancies. Chapter 3 details work on studying the differences in peptides and lipids in
healthy human placental explants maintained in an environment simulating proposed preeclamp-
tic conditions mainly hypoxia, oxidative stress and proinflammatory cytokines. In chapter 4 a
follow-up study was conducted to study all the altered biomolecules in placentae under
preeclamptic conditions in actual preeclamptic placentae and evaluate their statistical signifi-
cance by comparing them with controls.
36
1.8 References
1. Bell, M. J., A Historical Overview of Preeclampsia-Eclampsia. J. Obstet. Gynecol.
Neonatal Nurs. 2010, 39 (5), 510-518.
2. Adu-Bonsaffoh, K.; Antwi, D. A.; Obed, S. A.; Gyan, B., Nitric oxide dysregulation in the
pathogenesis of preeclampsia among Ghanaian women. Integr. Blood Press. Control 2015,
8, 1-6.
3. Singh, H. J., Pre-Eclampsia: Is It All In The Placenta? Malays. J. Med. Sci. 2009, 16 (1), 7-
15.
4. Brown, M. A.; Lindheimer, M. D.; de Swiet, M.; Assche, A. V.; Moutquin, J.-M., The
Classification and Diagnosis of the Hypertensive Disorders of Pregnancy: Statement from
the International Society for the Study of Hypertension in Pregnancy (ISSHP). Hypertens.
Pregnancy 2001, 20 (1), ix-xiv.
5. Osungbade, K. O.; Ige, O. K., Public health perspectives of preeclampsia in developing
countries: implication for health system strengthening. J. Pregnancy 2011, 2011, 481095.
6. Zhou, C. C.; Irani, R. A.; Dai, Y.; Blackwell, S. C.; Hicks, M. J.; Ramin, S. M.; Kellems,
R. E.; Xia, Y., Autoantibody-mediated IL-6-dependent endothelin-1 elevation underlies
pathogenesis in a mouse model of preeclampsia. J. Immunol. 2011, 186 (10), 6024-34.
7. English, F. A.; Kenny, L. C.; McCarthy, F. P., Risk factors and effective management of
preeclampsia. Integr. Blood Press. Control 2015, 8, 7-12.
8. Duley, L., The global impact of pre-eclampsia and eclampsia. Semin. Perinatol. 2009, 33
(3), 130-137.
9. Cincotta, R. B.; Brennecke, S. P., Family history of pre-eclampsia as a predictor for pre-
eclampsia in primigravidas. Int. J. Gynaecol. Obstet. 1998, 60 (1), 23-7.
37
10. Hutcheon, J. A.; Lisonkova, S.; Joseph, K. S., Epidemiology of pre-eclampsia and the
other hypertensive disorders of pregnancy. Best Pract. Res. Clin. Obstet. Gynaecol. 2011,
25 (4), 391-403.
11. Cartwright, J. E.; Fraser, R.; Leslie, K.; Wallace, A. E.; James, J. L., Remodelling at the
maternal–fetal interface: relevance to human pregnancy disorders. Reproduction 2010, 140
(6), 803-813.
12. Red-Horse, K.; Zhou, Y.; Genbacev, O.; Prakobphol, A.; Foulk, R.; McMaster, M.; Fisher,
S. J., Trophoblast differentiation during embryo implantation and formation of the
maternal-fetal interface. J. Clin. Invest. 2004, 114 (6), 744-754.
13. Huppertz, B.; Frank, H. G.; Reister, F.; Kingdom, J.; Korr, H.; Kaufmann, P., Apoptosis
cascade progresses during turnover of human trophoblast: analysis of villous
cytotrophoblast and syncytial fragments in vitro. Laboratory investigation; a journal of
technical methods and pathology 1999, 79 (12), 1687-1702.
14. Ashton, S. V.; Whitley, G. S.; Dash, P. R.; Wareing, M.; Crocker, I. P.; Baker, P. N.;
Cartwright, J. E., Uterine spiral artery remodeling involves endothelial apoptosis induced
by extravillous trophoblasts through Fas/FasL interactions. Arterioscler. Thromb. Vasc.
Biol. 2005, 25 (1), 102-8.
15. Young, B. C.; Levine, R. J.; Karumanchi, S. A., Pathogenesis of preeclampsia. Annu. Rev.
Pathol. 2010, 5, 173-92.
16. Uzan, J.; Carbonnel, M.; Piconne, O.; Asmar, R.; Ayoubi, J.-M., Pre-eclampsia:
pathophysiology, diagnosis, and management. Vasc. Health Risk Manage. 2011, 7, 467-
474.
38
17. Gilbert, J. S.; Ryan, M. J.; LaMarca, B. B.; Sedeek, M.; Murphy, S. R.; Granger, J. P.,
Pathophysiology of hypertension during preeclampsia: linking placental ischemia with
endothelial dysfunction. Am. J. Physiol. 2008, 294 (2), H541-H550.
18. Rajakumar, A.; Whitelock, K. A.; Weissfeld, L. A.; Daftary, A. R.; Markovic, N.; Conrad,
K. P., Selective overexpression of the hypoxia-inducible transcription factor, HIF-2alpha,
in placentas from women with preeclampsia. Biol. Reprod. 2001, 64 (2), 499-506.
19. Roberts, J. M.; Escudero, C., The placenta in preeclampsia. Pregnancy Hypertens. 2012, 2
(2), 72-83.
20. Sánchez-Aranguren, L. C.; Prada, C. E.; Riaño-Medina, C. E.; Lopez, M., Endothelial
dysfunction and preeclampsia: role of oxidative stress. Front. Physiol. 2014, 5, 372.
21. Myatt, L., Role of placenta in preeclampsia. Endocrine 2002, 19 (1), 103-11.
22. Hung, T. H.; Skepper, J. N.; Burton, G. J., In vitro ischemia-reperfusion injury in term
human placenta as a model for oxidative stress in pathological pregnancies. Am. J. Pathol.
2001, 159 (3), 1031-43.
23. Goel, A.; Rana, S., Angiogenic factors in preeclampsia: potential for diagnosis and
treatment. Curr. Opin. Nephrol. Hypertens. 2013, 22 (6), 643-650.
24. Woolcock, J.; Hennessy, A.; Xu, B.; Thornton, C.; Tooher, J.; Makris, A.; Ogle, R.,
Soluble Flt-1 as a diagnostic marker of pre-eclampsia. Aust. N. Z. J. Obstet. Gynaecol.
2008, 48 (1), 64-70.
25. Wang, A.; Rana, S.; Karumanchi, S. A., Preeclampsia: the role of angiogenic factors in its
pathogenesis. Physiology (Bethesda) 2009, 24, 147-58.
39
26. Vitoratos, N.; Economou, E.; Iavazzo, C.; Panoulis, K.; Creatsas, G., Maternal Serum
Levels of TNF-Alpha and IL-6 Long after Delivery in Preeclamptic and Normotensive
Pregnant Women. Mediators Inflamm. 2010, 2010.
27. Saito, S.; Sakai, M., Th1/Th2 balance in preeclampsia. J. Reprod. Immunol. 2003, 59 (2),
161-73.
28. Conrad, K. P.; Benyo, D. F., Placental cytokines and the pathogenesis of preeclampsia.
Am. J. Reprod. Immunol. 1997, 37 (3), 240-9.
29. Konermann, L.; Ahadi, E.; Rodriguez, A. D.; Vahidi, S., Unraveling the Mechanism of
Electrospray Ionization. Anal. Chem. 2013, 85 (1), 2-9.
30. Banerjee, S.; Mazumdar, S., Electrospray Ionization Mass Spectrometry: A Technique to
Access the Information beyond the Molecular Weight of the Analyte. Int. J. Anal. Chem.
2012, 2012, 40.
31. Kebarle, P.; Verkerk, U. H., Electrospray: from ions in solution to ions in the gas phase,
what we know now. Mass Spectrom. Rev. 2009, 28 (6), 898-917.
32. Croxatto, A.; Prod'hom, G.; Greub, G., Applications of MALDI-TOF mass spectrometry in
clinical diagnostic microbiology. FEMS Microbiol. Rev. 2012, 36 (2), 380-407.
33. Chernushevich, I. V.; Loboda, A. V.; Thomson, B. A., An introduction to quadrupole-
time-of-flight mass spectrometry. J. Mass Spectrom. 2001, 36 (8), 849-65.
34. Vestal, M. L.; Juhasz, P.; Martin, S. A., Delayed extraction matrix-assisted laser desorption
time-of-flight mass spectrometry. Rapid Commun. Mass Spectrom. 1995, 9 (11), 1044-
1050.
35. Cotter, R. J.; Gardner, B. D.; Iltchenko, S.; English, R. D., Tandem time-of-flight mass
spectrometry with a curved field reflectron. Anal. Chem. 2004, 76 (7), 1976-81.
40
36. Graves, P. R.; Haystead, T. A., Molecular biologist's guide to proteomics. Microbiol. Mol.
Biol. Rev. 2002, 66 (1), 39-63; table of contents.
37. Vogel, C.; Marcotte, E. M., Insights into the regulation of protein abundance from
proteomic and transcriptomic analyses. Nature reviews. Genetics 2012, 13 (4), 227-232.
38. Rozek, W.; Ciborowski, P. S., Proteomics and Genomics. In Neuroimmune Pharmacology,
Gendelman, H. E.; Ikezu, T., Eds. Springer US: Boston, MA, 2008; pp 725-741.
39. Hortin, G. L.; Sviridov, D., The dynamic range problem in the analysis of the plasma
proteome. J. Proteomics 2010, 73 (3), 629-636.
40. Gundry, R. L.; White, M. Y.; Murray, C. I.; Kane, L. A.; Fu, Q.; Stanley, B. A.; Van Eyk,
J. E., Preparation of Proteins and Peptides for Mass Spectrometry Analysis in a Bottom-Up
Proteomics Workflow. Curr. Protoc. Mol. Biol. 2009, CHAPTER, Unit10.25-Unit10.25.
41. Kedia, K.; Nichols, C. A.; Thulin, C. D.; Graves, S. W., Novel "omics" approach for study
of low-abundance, low-molecular-weight components of a complex biological tissue:
regional differences between chorionic and basal plates of the human placenta. Anal.
Bioanal. Chem. 2015, 407 (28), 8543-56.
42. Ma, B.; Zhang, K.; Hendrie, C.; Liang, C.; Li, M.; Doherty-Kirby, A.; Lajoie, G., PEAKS:
powerful software for peptide de novo sequencing by tandem mass spectrometry. Rapid
Commun. Mass Spectrom. 2003, 17 (20), 2337-42.
43. Zhao, Y.-Y.; Wu, S.-P.; Liu, S.; Zhang, Y.; Lin, R.-C., Ultra-performance liquid
chromatography–mass spectrometry as a sensitive and powerful technology in lipidomic
applications. Chem. Biol. Interact. 2014, 220, 181-192.
41
44. Köfeler, H. C.; Fauland, A.; Rechberger, G. N.; Trötzmüller, M., Mass Spectrometry
Based Lipidomics: An Overview of Technological Platforms. Metabolites 2012, 2 (1), 19-
38.
45. Cajka, T.; Fiehn, O., Comprehensive analysis of lipids in biological systems by liquid
chromatography-mass spectrometry. TrAC Trends Anal. Chem. 2014, 61, 192-206.
46. Alvarez, M. T.; Shah, D. J.; Thulin, C. D.; Graves, S. W., Tissue proteomics of the low-
molecular weight proteome using an integrated cLC-ESI-QTOFMS approach. Proteomics
2013, 13 (9), 1400-11.
47. Banks, R. E.; Dunn, M. J.; Forbes, M. A.; Stanley, A.; Pappin, D.; Naven, T.; Gough, M.;
Harnden, P.; Selby, P. J., The potential use of laser capture microdissection to selectively
obtain distinct populations of cells for proteomic analysis--preliminary findings.
Electrophoresis 1999, 20 (4-5), 689-700.
48. Walch, A.; Rauser, S.; Deininger, S.-O.; Höfler, H., MALDI imaging mass spectrometry
for direct tissue analysis: a new frontier for molecular histology. Histochem. Cell Biol.
2008, 130 (3), 421-434.
49. Berg, D.; Malinowsky, K.; Reischauer, B.; Wolff, C.; Becker, K. F., Use of formalin-fixed
and paraffin-embedded tissues for diagnosis and therapy in routine clinical settings.
Methods Mol. Biol. 2011, 785, 109-22.
50. Centlow, M.; Hansson, S. R.; Welinder, C., Differential proteome analysis of the
preeclamptic placenta using optimized protein extraction. Biomed. Res. Int. 2009, 2010.
51. Robinson, J. M.; Ackerman, W. E. t.; Kniss, D. A.; Takizawa, T.; Vandre, D. D.,
Proteomics of the human placenta: promises and realities. Placenta 2008, 29 (2), 135-43.
42
52. Centlow, M.; Hansson, S. R.; Welinder, C., Differential Proteome Analysis of the
Preeclamptic Placenta Using Optimized Protein Extraction. J. Biomed. Biotechnol. 2010,
2010, 9.
53. Mine, K.; Katayama, A.; Matsumura, T.; Nishino, T.; Kuwabara, Y.; Ishikawa, G.; Murata,
T.; Sawa, R.; Otsubo, Y.; Shin, S.; Takeshita, T., Proteome analysis of human placentae:
pre-eclampsia versus normal pregnancy. Placenta 2007, 28 (7), 676-87.
54. Wang, F.; Shi, Z.; Wang, P.; You, W.; Liang, G., Comparative Proteome Profile of Human
Placenta from Normal and Preeclamptic Pregnancies. PLoS One 2013, 8 (10), e78025.
55. Vandre, D. D.; Ackerman, W. E. t.; Tewari, A.; Kniss, D. A.; Robinson, J. M., A placental
sub-proteome: the apical plasma membrane of the syncytiotrophoblast. Placenta 2012, 33
(3), 207-13.
56. Epiney, M.; Ribaux, P.; Arboit, P.; Irion, O.; Cohen, M., Comparative analysis of secreted
proteins from normal and preeclamptic trophoblastic cells using proteomic approaches. J.
Proteomics 2012, 75 (6), 1771-7.
57. Ma, K.; Jin, H.; Hu, R.; Xiong, Y.; Zhou, S.; Ting, P.; Cheng, Y.; Yang, Y.; Yang, P.; Li,
X., A proteomic analysis of placental trophoblastic cells in preeclampsia-eclampsia. Cell
Biochem. Biophys. 2014, 69 (2), 247-58.
58. Korkes, H. A.; Sass, N.; Moron, A. F.; Câmara, N. O. S.; Bonetti, T.; Cerdeira, A. S.; Da
Silva, I. D. C. G.; De Oliveira, L., Lipidomic Assessment of Plasma and Placenta of
Women with Early-Onset Preeclampsia. PLoS One 2014, 9 (10), e110747.
59. Baig, S.; Lim, J. Y.; Fernandis, A. Z.; Wenk, M. R.; Kale, A.; Su, L. L.; Biswas, A.;
Vasoo, S.; Shui, G.; Choolani, M., Lipidomic analysis of human placental
43
syncytiotrophoblast microvesicles in adverse pregnancy outcomes. Placenta 2013, 34 (5),
436-42.
60. Tu, C.; Rudnick, P. A.; Martinez, M. Y.; Cheek, K. L.; Stein, S. E.; Slebos, R. J. C.;
Liebler, D. C., Depletion of Abundant Plasma Proteins and Limitations of Plasma
Proteomics. J. Proteome Res. 2010, 9 (10), 4982-4991.
61. Millioni, R.; Tolin, S.; Puricelli, L.; Sbrignadello, S.; Fadini, G. P.; Tessari, P.; Arrigoni,
G., High abundance proteins depletion vs low abundance proteins enrichment: comparison
of methods to reduce the plasma proteome complexity. PLoS One 2011, 6 (5), e19603.
62. Tirumalai, R. S.; Chan, K. C.; Prieto, D. A.; Issaq, H. J.; Conrads, T. P.; Veenstra, T. D.,
Characterization of the low molecular weight human serum proteome. Mol. Cell.
Proteomics 2003, 2 (10), 1096-103.
63. Merrell, K.; Southwick, K.; Graves, S. W.; Esplin, M. S.; Lewis, N. E.; Thulin, C. D.,
Analysis of low-abundance, low-molecular-weight serum proteins using mass
spectrometry. J Biomol Tech 2004, 15 (4), 238-48.
64. Bäckvall, H.; Lehtiö, J. The Low Molecular Weight Proteome: Methods and Protocols,;
Springer: New York, 2013
44
2 Novel ‘omics’ approach to study low abundance, low molecular weight components in
a complex biological tissue: Regional differences between chorionic and basal plates of
human placenta
This chapter in large part was reproduced with permission from a published research article:
Kedia, K.; Nichols, C. A.; Thulin, C. D.; Graves, S. W., Novel "omics" approach for study of low-
abundance, low-molecular-weight components of a complex biological tissue: regional differences
between chorionic and basal plates of the human placenta. Analytical and bioanalytical chemistry
2015, 407 (28), 8543-56.
2.1 Abstract
Tissue proteomics has relied heavily on two-dimensional gel electrophoresis for protein
separation and quantitation, followed by single protein isolation, trypsin digestion and mass
spectrometric protein identification. Such methods observe predominantly higher abundance, full
length proteins. Tissue peptidomics has been recently developed but still interrogates the more
highly abundant species, resulting often in only dozens of peptides being observed and identi-
fied. Tissue lipidomics is likewise new and reported studies limited. We have developed an ‘om-
ics’ approach that surveys over 7,000 endogenous, lower molecular weight, low abundance spe-
cies and have applied this to a human tissue, placenta. Placenta is thought to play a role in com-
plications of pregnancy and its proteomic evaluation is of substantial interest. Past research on
the placental proteome has studied abundant, higher molecular weight proteins. Large scale,
global proteomics or peptidomics have had limited application to placenta but either would be
challenging owing to the anatomic complexity and broad concentration range of proteins in this
45
tissue. Our approach, involving protein depletion, capillary liquid chromatography, and tandem
mass spectrometry attempted to identify molecular differences between two placental regions of
the same tissue having only slightly different cellular composition. Our analysis revealed 16 spe-
cies with statistically significant differences between the two regions. Tandem mass spectrome-
try successfully sequenced or otherwise identified twelve of these. The method’s successfully
finding and identifying regional differences in the expression of low-abundance, lower molecular
weight biomolecules demonstrates the potential of our approach.
2.2 Introduction
Tissue proteomics has historically relied on single- or two-dimensional gel electrophore-
sis (2-DGE) for separation and quantitation of tissue proteins. This has typically been followed
by single protein isolation, trypsin digestion and mass spectrometric (MS) analysis to provide
protein identification by database comparison.1, 2 These approaches survey highly abundant,
larger molecular weight proteins. These methods, without tandem MS fragmentation studies em-
ploying collisional-induced dissociations or electron-transfer dissociation, do not provide direct
amino acid sequences. None of these methods, in the absence of fragmentation, identifies post-
translational modifications (PTMs) or chemical changes that can occur in response to reactive
oxygen species (ROS) or chemical modifiers directly. Other ‘global’ proteomic approaches, such
as multi-dimensional protein identification technology (MudPIT) rely on successive chromato-
graphic separations, most often after trypsin digestion, rather than electrophoretic methods.3
They have also been used for the assessment of tissues.4 These global approaches often involve
amino acid sequencing of tryptic digest fragments to identify proteins and associated PTMs.
Such global methods survey a greater number of full length proteins than 2-DGE but generally
46
still suffer from ion suppression and masking of low-abundance components. They are not cur-
rently designed to study peptides or lipids. More recently, top-down methods have been devel-
oped that allow for the injection of intact, full-length proteins onto the MS system and can in-
clude amino acid sequencing and the observation of PTMs.5, 6 However, such approaches often
require more than one separation step prior to MS analysis, survey proteins that have higher
charge states yielding sufficiently low mass to charge ratios and typically require very highly
mass-accurate instruments such as Fourier transform ion cyclotron resonance mass spectrome-
ters. While proteomics studies are valuable, they still focus on higher molecular weight, more
highly abundant proteins. The highly expressed cell proteins typically measured are often cyto-
skeletal elements or are involved in protein folding and trafficking or perform various house-
keeping functions and are less frequently involved in regulation and disease processes. Hence,
interrogating smaller, lower abundance components in cells may be as important as probing the
higher abundance proteins.
Alternatively, tissue peptidomics has been recently developed and employed for cata-
loguing peptides present in a particular tissue or for quantitative comparisons between tissues us-
ing both label free7 and isotope labeled tags (ICAT, iTRAQ).8, 9 Labelling techniques such as
iTRAQ applied to tissue peptides appears to measure dozens of peptides.9 However, different la-
beling agents react inconsistently with different peptides suggesting variable quantitation.9 Such
techniques are not useful for lipids. These peptidomics methods hold promise but appear limited
in the number and type of small cell constituents assayed. Tissue lipidomics frequently employs
tissue homogenization followed by organic solvent extraction and injection of extracted mole-
cules onto to the MS without or with prior chromatographic separation.10-12 Fragmentation stud-
ies have been attempted but there is not a well-developed reference database to allow for lipid
47
identification from fragment mass comparison. Identification requires careful assessment of mass
differences between individual fragment peaks to characterize the neutral loss components and
thereby identify molecular parts present in the pre-fragmentation parent species.
We have developed a unique and complementary approach to analyze low abundance,
lower molecular weight components in tissue specimens.13 We employ a protein depletion step.14
This desorbs low molecular weight species from binding partners and with the removal of highly
abundant proteins allows for the interrogation of several thousand additional molecular species.
The method has been applied to tissue as part of an initial study evaluating its analytical reliabil-
ity and its breadth of coverage.13 It was sufficiently reproducible to allow for label free quantita-
tive comparisons between different organs and was able to observe over 7,000 molecular spe-
cies.13 This method considers lower abundance, small proteins, peptides, lipids and potentially
other biomolecules and allows for relatively seamless tandem MS fragmentation studies to pro-
vide de novo amino acid sequencing for polypeptides and chemical characterization and classifi-
cation of lipids.15 Such fragmentation data can identify post-translational modifications, oxida-
tive damage and/or other non-physiologic chemical modifications. In this report this novel ap-
proach has been applied to human tissue.
Certainly, one of the most challenging but important targets for tissue proteomics is the
human placenta. The placenta is of clinical interest because of its almost certain involvement in
preeclampsia (PE), a disease complicating 2-10% of all pregnancies that results in as many as
75,000 maternal deaths yearly.16 The placenta is required for PE and evidence suggests altera-
tions in placental biochemistry occur prior to and likely lead to established PE.17
48
One rather unique challenge in working with the placenta is that it is formed from cells
originating from both the fetus and the mother but the cells are interspersed rather than compart-
mentalized (Figure 2.1).44 Emerging evidence suggests that specific early changes that may con-
tribute to PE arise sometimes in the fetal and sometimes in the maternal derived tissues.18 These
studies have usually measured single protein targets, but nonetheless underscore the importance
of distinguishing between fetal and maternal regions in the placenta if one is to understand PE
fully.19 Fixing and staining tissue may allow a particular cell type to be identified and immuno-
staining may allow for monitoring of a specific protein’s abundance in that cell type but this ap-
proach is not amenable to global analysis of molecular mediators or modified dynamics. Further-
more, quantitative comparative proteomic approaches would be difficult, if not impossible, to
carry out after such chemical modifications even though there are methods that can reverse much
of the cross-linking initiated by fixatives.20 Placental proteomics has only been attempted rela-
tively recently and has mostly studied highly abundant proteins.21, 22 There have been a small
number of more global proteomics studies attempted on the placenta.23, 24 These studies have fo-
cused on a single class of biomolecules, for example proteins23 or metabolites.24 However, there
do not appear to have been other ‘omics’ studies to probe the lower molecular weight compo-
nents of placenta. Additionally, there has been no effort in prior placental proteomic studies to
investigate regional differences within the placenta. Consequently, we considered the use of our
novel tissue ’omics’ approach for this target as an important demonstration of the method’s util-
ity. The intent of this study was to assess the robustness of the method in finding region-specific,
quantitative differences in molecular components, but not to develop a set of cell-specific bi-
omarkers. If this approach is successful, it provides a method for the exploration of thousands of
previously unstudied, smaller molecular weight, low abundance components in complex tissues.
49
It also provides more particularly a way of exploring events potentially mediating diseases, in-
cluding PE, where one would anticipate that many more differences would be seen between pla-
cental regions as part of disease and hence, this approach may allow for clarification not only of
which molecules are modified but more specifically clarify whether they have a fetal versus a
maternal origin.
2.3 Materials and methods
2.3.1 Tissue collection
Internal Review Boards of Intermountain Health Care and its affiliated hospitals and of
Brigham Young University provided approval of these studies. Placentas were considered dis-
carded materials and no informed consent was obtained, and no personal information was pro-
vided. Placentas were collected immediately following uncomplicated vaginal or C-section de-
liveries (n=12). The fetal amnionic membrane was carefully removed and thin tissue pieces were
dissected from the extreme outer surface of the chorionic plate. Placentas were then inverted and
on the opposing surface (basal plate), thin tissue pieces were sliced from the very outermost sur-
face of the cotyledons. Both tissue types were collected from a location midway between the
cord and the peripheral edge of the placenta along the same vertical axis. Care was taken to
maintain consistency in this process. After collection, tissues were rapidly frozen on dry ice and
maintained frozen until processed further.
50
2.3.2 Tissue homogenization
Frozen placental tissue from one location was finely minced using a sterile stainless steel
blade. These shavings (around 500 mg) were then placed in a 5 mL stainless steel ball mill cylin-
der containing 12–15, 3-mm chromium steel grinding balls. The vessel was flash frozen in liquid
nitrogen for 4 min and shaken using a ball mill dismembrator (Mikro-Dismembrator S, Sartorius,
Göttingen, Germany) operating at 2000 rpm for 8-10 minutes to form a paste-like homogenate.
The tissue homogenate was re-suspended in 5 mL of cold physiological phosphate buffered sa-
line (PBS, pH 7.2). This mixture was vortexed thoroughly and centrifuged (Sorval RT7, Kendro
Laboratory Products, Newton, CT) for 10 min at 4000 rpm (8046xg) at 4°C. Aliquots of 200 µL
of the supernatant were stored at -80°C for further processing. A broad-spectrum proteolytic en-
zyme inhibitor cocktail was added to specimen prior to processing (Sigma Aldrich, catalogue
number P9599). Additionally, processing at cold temperatures and immediate freezing were em-
ployed to further prevent proteolysis.
2.3.3 Protein depletion
Highly abundant, high molecular weight proteins were precipitated by adding 2:1 (v/v)
acetonitrile to the homogenate. The mixture was vortexed briefly, placed at room temperature for
30 min, and centrifuged (IEC Mcromax RF, Thermo Fisher Scientific, Waltham, MA) at 14,000
rpm (13107xg) for 10 min at 4°C. The pellet was discarded and the supernatant (~550 μL) was
transferred to a new Eppendorf tube containing 300 μL of HPLC-grade water. The total volume
was then reduced to 200 μL using a vacuum centrifuge (CentriVap Concentrator, Labconco Cor-
poration, Kansas City, MO) at 37°C. The apparent protein concentration of each sample was de-
51
termined (Pierce Microplate BCA Protein Assay Kit, Thermo Scientific, Waltham, MA). A vol-
ume of the lyophilized sample equivalent to 20 μg of protein was then dried in vacuo to a volume
of 10 μL. Then 10 μL of 88% formic acid was added, each sample briefly vortexed, and 5 µL of
sample having a concentration of 1 μg/μL apparent protein injected onto the capillary liquid
chromatography-mass spectrometer (cLC/MS) system. This single protein-depleted specimen
contained small proteins, peptides, lipids and potentially other metabolites.
2.3.4 Chromatographic separation
Reverse-phase capillary liquid chromatography (cLC) separation was then performed us-
ing an LC Packings Ultimate Capillary HPLC pump system, with a FAMOS® autosampler (Di-
onex Corporation, Sunnyvale, CA, USA). The system involved a 1 mm (16.2 µl) dry-packed Mi-
croBore guard column (IDEX Health and Science, Oak Harbor, WA), coupled to a 15 cm x 250
µm i.d. capillary column, slurry-packed in-house with POROS R1 reverse-phase medium (Ap-
plied Biosystems, Foster City, CA). The elution gradient was generated using an aqueous phase:
98% HPLC grade water, 2% acetonitrile, 0.1% formic acid and an organic phase: 98% acetoni-
trile, 2% HPLC grade water, 0.1% formic acid. This packing material had been previously used
over a prolonged period for experiments involving both serum and tissue proteomics and allowed
continuity in experiments with good chromatographic reproducibility.14 The gradient began with
3 min of 95% aqueous and 5% organic phase, followed by a linear rise in organic phase to 60%
over the next 24 min. The gradient was then increased linearly to 95% organic phase/5% aqueous
phase over the next 7 min, held at 95% organic phase for 7 min and returned to 95% aqueous
phase over 5 min. The column was allowed to re-equilibrate until the end time of the run (58
min). The flow rate was 5 µL/min.
52
2.3.5 Mass spectrometric analysis
Eluate was introduced through an electrospray needle into a tandem mass spectrometer
(QSTAR Pulsar І quadrupole orthogonal time-of-flight, QTOF, Applied Biosystems, Foster City,
CA). The electrospray source was at 4800 V. All samples were run in the positive ion mode.
Other MS parameters used during the analyses were: microchannel plate voltage (MCP): 2380 V,
plate voltage: 330 V, curtain gas pressure: 20 psi, source gas pressure: 20 psi. A mass spectrum
was obtained every 1 sec from m/z 500–2500 over 5–55 minutes elution. Fragmentation studies
were carried out with an acquisition rate of 1 spectrum/second. MS calibration was performed by
injection of a non-physiological peptide, Glu-1-Fibrinopeptide (GluFib) having an m/z of 785.85
(z=+2). Resolution of the instrument during the time when these samples were run varied be-
tween 9,500-10,000. The results were analyzed using Analyst QS® 1.1 software (Applied Bio-
systems, Foster City, CA, USA). Based on tissue homogenates spiked with this same calibrating
peptide, the approach is estimated to have a limit of detection (LOD) of 5-10 nM with the great
majority of species interrogated having estimated concentrations of 7-200 nM, representing as
little as 2 pg of material per MS peak.
2.3.6 Time normalization and data analysis
Elution times can vary from day to day and occasionally from run to run. To compensate,
we defined 2 min windows within the useful chromatogram. In each window we selected a cen-
tral endogenous peak representing a molecule found in all specimens.
These central peaks eluted at approximately 2 min intervals. These were used for time
alignment among different runs. To find regional tissue markers, mass spectra from chorionic
plate and basal plate were color coded, overlaid, centered on a time marker and peaks visually
53
inspected to find features appearing to differ in abundance. Every peak appearing to demonstrate
quantitative differences between comparison groups was further evaluated. This process in-
volved ‘extracting’ the candidate marker using Analyst software and recording the maximum
height of the XIC (extracted ion chromatogram) peak. Time markers selected to correct for varia-
tion in elution times (in order of increasing elution times) had mass-to-charge ratios as follow:
695.12 (z=+4), 827.77 (z=+6), 685.9 (z=+4), 672.7 (z=+3), 1009.4 (z < +10), 616.19 (z=+1),
526.3 (z=+1), 524.3 (z=+1), 650.4 (z=+1), 675.56 (z=+1). The peptide m/z 1009.4 had a charge
estimated as a +14.
The analytical performance of the method has been previously evaluated 13 and demon-
strated coefficients of variation from 23.0 to 26.8% for single analytes being measured in differ-
ent precipitates of the same tissue. This compares favorably to proteomics results using 2-DGE-
MS methods.25, 26 Other global or top-down proteomics methods have not frequently reported
quantitative reproducibility, but one study reported a CV of 36% for technical replicates.27
2.3.7 Statistical analysis
A Student’s t-test was performed to determine whether the abundance of an extracted
peak was different between the two regions. A p-value of less than 0.05 was considered statisti-
cally significant. Additionally, the consistency of these differentially expressed molecules to be
increased in the chorionic or basal plates was assessed by constructing 2 x 2 contingency tables
and applying a Fisher Exact test to the results (Table 2.1). Finally, as an indication of biologic
variability, we have calculated relative standard deviations (%RSD) for each of the 16 differen-
tially expressed species in each region and included these in Table 2.1.
54
2.3.8 Normalization of candidate markers to reduce non-biological variation
For all statistically significant candidate markers a reference peak was selected. This was
an endogenous peak that eluted in a time interval closer to the candidate marker and was present
in all samples in comparable abundance across comparison groups. To normalize candidate
peaks, the intensity of the candidate marker was divided by the intensity of the reference peak.
Reference peaks selected to normalize species in elution window 1, 2, 3, 4, 6, 7 and 10 had the
following mass to charge ratios: 1241.17 (z=+4), 993.12 (z=+5), 660.38 (z=+1), 660.38 (z=+1),
913.48 (z=+1), 913.48 (z=+1), and 563.57 (z=+1). All these reference peaks were quantitatively
the same between chorionic and basal plates (p > 0.1). This approach has been previously evalu-
ated with an overall coefficient of variation of ~25% in terms of combined pre-analytical and an-
alytical variability.13 A Student’s t-test was then performed a second time on normalized values.
2.3.9 Identification of peptides of interest by tandem MS
Significantly different molecules were submitted to tandem MS with collisionally-in-
duced dissociation (CID). The fragmentation pattern produced was used in an attempt to deter-
mine the amino acid sequence and parent protein. One or more samples containing high levels of
the peptide of interest were thawed and 5 µg of the sample containing 2.5 µL of 88% formic acid
was hand-injected into the cLC-MS system. Fragmentation data were collected for 2 minutes us-
ing argon or nitrogen as the collision gas. Different collision energies were used to provide more
complete production of b- and y-series fragment ions. Spectra obtained from different collision
energies were summed to obtain more complete coverage. Charge states of peaks in the export
data list were checked for any misassigned charge states. Charge states for all the peaks were
converted to their +1 m/z values by using the formula +1 mass = m/z value * charge – (charge –
55
1H+). This corrected list was then submitted to the search engine Mascot (Mascot 2.3, www.ma-
trixscience.com) which correlated the fragmentation data with archived protein sequence data.
2.3.10 De novo sequencing
De novo sequencing was attempted on peptides when Mascot failed.28 Individual amino
acids were assigned by measuring the mass differences between fragment ions. Presence of im-
monium ions aided in identifying the presence of particular amino acids. This process proved to
be much more complicated because these peptides were not produced by trypsin digestion. A
BLAST search using the National Institutes of Health website (http://blast.ncbi.nlm.nih.gov) was
performed on the sequence of specific amino acids observed from either Mascot assignment or
de novo sequencing and the calculated probability (an E score) of having identified the parent
protein provided in Table 2.2
2.3.11 Identification of lipid markers
The processing step, involving acetonitrile treatment, produced a specimen that contained
in addition to polypeptides, polar or positively charges lipids, including neutral lipids that formed
adducts with cations in the elution buffer. During the cLC step, these lipids, mainly phosphati-
dylcholines (PC) and sphingomyelins (SM) eluted typically at mobile phase organic solvent con-
centrations of 40% or greater. These lipids become positively charged when sprayed in an acidic
mobile phase due to the presence of a quarternary nitrogen in the choline group and protonation
of the phosphate group and enter the mass spectrometer, most often as the [M+H]+ species. Li-
pids were characterized by the same tandem MS-MS approach as used for peptides with the
same instrument parameters. The collision gas was usually nitrogen, and 1 or 2 collision energies
56
were sufficient to fragment the lipid of interest. LIPID MAPS (lipidmaps.org) database was used
to investigate the structure based on parent mass and molecular formula of the concerned lipid,
while LIPID MS predictor (http://www.lipidmaps.org/tools/index.html) was used to calculate the
possible fragment ions.
2.4 Results
2.4.1 Time markers to align chromatographic elution times
As described above, the total ion chromatogram (TIC) was divided into 2 min windows
using time alignment markers. These time windows covered most of the useful chromatogram
with its attendant TIC. The many mass spectra recorded during the 2 min were averaged. The
time markers are listed in the Methods section above.
2.4.2 Differentially abundant low abundance, LMW weight biomolecules between chorionic
plate and basal plate.
MS data was color coded by region and overlays of MS data for a given time window were
visually inspected to find obvious differences in peak abundance between comparison groups (at
least 1.5x or greater). As expected, most peaks were common to both sides of the placenta and
present in comparable abundance. Importantly, however a number of peaks differed significantly
in their intensity between chorionic plate and basal plate tissue, indicating regional selectivity.
Figure 2.2 provides an example. The results demonstrated 16 peaks that exhibited statistically
different abundances between the two sides. Table 2.1 summarizes the MS characteristics of
these 16 peaks.
57
Figure 2.1. An overlay of 16 mass spectra in the region containing peptide m/z 718.36 with its isotope envelope (z = +2). Red: tissue collected from chorionic plate; blue: tis-sue collected from outer cotyledons (basal plate). This species was more abundant in the chorionic plate of placenta (p=0.007).
58
Table 2.1. List of 16 differentially molecular species across all 10 time windows.
Marker(m/z)
z
Student’s t-test Fisher exact
%RSD (Raw)
%RSD (Normalized)
Raw p-value
Normalized p- value
p-value
Basal plate
Chorionic plate
Basal plate
Chorionic plate
Window 1 (XIC range: m/z 695-696)
624.32 +2 0.02 0.02 0.0033 31.7% 32.5% 44.7% 45.8% 718.36 +2 0.005 0.007 0.0001 54.9% 47.6% 68.1% 58.1%
Window 2 (XIC range: m/z 827-828)
614.38 +3 0.05 0.02 0.0033 41.4% 41.8% 36.7% 29.5% 808.87 +2 0.05 0.05 0.0033 54.8% 49.9% 58.4% 45.7%
Window 3 (XIC range: 685-686)
760.4 +4 0.06 0.01 0.039 33.0% 38.2% 29.0% 34.0% Window 4 (XIC win-
dow: 672-673)
857.15 +10 0.01 0.02 0.22 17.3% 18.1% 28.8% 30.0% Window 6 (XIC range:
616-617)
520.33 1 0.02 0.03 0.039 42.3% 41.6% 41.9% 41.7% 542.31 1 0.01 0.03 0.039 34.1% 38.9% 38.9% 38.1% 544.33 1 0.01 0.01 0.039 46.2% 45.1% 46.1% 44.9% 566.32 1 0.02 0.03 0.039 39.0% 21.9% 38.9% 21.5% 568.33 1 0.03 0.04 0.039 56.6% 44.7% 54.1% 44.6%
Window 7 (XIC range: 526-527)
570.34 1 0.03 0.04 0.22 50.4% 36.4% 48.9% 35.8% 1159.50 1 0.02 0.02 0.0001 33.2% 20.4% 31.5% 15.7%
Window 10 (XIC range: 675-676)
647.55 1 0.07 0.01 0.039 36.5% 39.9% 18.1% 22.5% 673.52 1 0.01 0.001 0.0033 29.8% 37.5% 20.9% 23.4% 701.55 1 0.09 0.01 0.039 19.5% 27.9% 23.6% 24.0%
59
2.4.3 Identification of peptides
Tandem mass spectrometry (MS/MS) was performed to determine amino acid sequence
of significantly different regional peptide markers. Tandem MS/MS was also used to fragment
candidate lipid markers, allowing the lipid class to be established.
Of the 7 peptide peaks, two represented the same molecular species observed in two dif-
ferent charge states. Of the 6 unique peptide biomarkers, Mascot provided sequences for two.
For example, the peptide m/z 718.36 (z = +2) was identified as fetal Hgb (γ subunit).
Two peptides were identified by de novo sequencing. The marker with m/z of 808.82
(z=+2) and a neutral mass of 1615.64 Da was identified as fibrinopeptide A phosphorylated at
Ser-3. We were able to determine 8 amino acid in series from the MS/MS spectrum. A BLAST
search was performed using the NIH protein database to identify the parent protein as fibrinopep-
tide A. This peptide has a predicted neutral mass of 1535.68 Da, which differed from the parent
mass of the marker by ~80 Da indicating the presence of a phosphate group. The sequence of this
peptide of fibrinopeptide A was entered into a fragment ion calculator to predict individual b and
y ions. A value of 80 was added to the serine residue, as the site for phosphorylation (given the
absence of threonine and tyrosine, ADSGEGDFLAEGGGVR) which gave good concordance
between predicted and observed fragments. This confirmed the identity of this peptide as a phos-
phorylated fragment fibrinopeptide A (Figure 2.3).
Additionally, a peptide with m/z of 760.4 and charge state of +4 was identified by de
novo sequencing. We identified 4 amino acids in series and performed a BLAST search. The par-
ent protein for this peptide was assigned as hemoglobin (Hgb) alpha-2 based on sequence homol-
ogy and the ability to achieve the observed overall mass of the peptide within the sequence of the
parent protein.
60
Two peptides m/z 857.49 and 624.32 could not be sequenced due to their high charge
state and low abundance. The results for all peptide identities are summarized in Table 2.3.
61
Table 2.2. Summary of the chemical identities of peptides differentially expressed in chorionic plate and basal plate of human placenta.
Peptide marker (m/z)
z Neutral mass
Parent protein Amino acid sequenceb E-score
718.36 +2 1434.72 Hgba subunit gamma (Fetal Hgb)
GHFTEEDKATITS 5E-07
614.38 +3 1840.14 Hgb subunit alpha-1
VLSPADKTNVKAAWGKVG 8E-11
808.82 +2 1615.64 Phosphorylated fi-brinopeptide A
ADpS*GEGDFLAEGGGVR 0.013
760.68 +4 3038.72 Hgb subunit alpha-2
VLSPADKTNVKAAWGKVGAHAG-EYGAEALE
3.3
aHgb: Hemoglobin, bSequence represents the amino acid composition of the peptide identified, *pS in the sequence of 808.82 represents phosphorylation at amino acid serine.
62
Figure 2.2. Averaged MS2 spectra of a second differentially expressed peptide m/z 808.82 (z = +2) (mass accuracy of the precursor peak =27 ppm) and the MS fragment-ion calculator’s prediction of b and y ions. This peptide was identified as a fragment of phosphorylated fibrinopeptide A.
63
2.4.4 Identification of lipids
There were 10 quantitatively different, region-specific markers likely to be lipids given
their elution later in the cLC chromatogram consistent with their being non-polar and their frag-
mentation patterns (See Table 2.3). Fragmentation spectra demonstrated an intense peak at m/z
184.07 characteristic of a phosphocholine group found in both phosphatidylcholines (PC) and
sphingomyelins (SM). An accurately determined mass for each candidate lipid was entered into
the search function of LIPID MAPS database, which presented possible structures and molecular
formulas for lipids, hence differentiating PC from SM. This software, however, gives no infor-
mation regarding likely tandem MS fragment ions. In order to further characterize the markers,
LIPID MS predictor was used, which predicts in silico possible fragments with their masses. If
these m/z values are then observed in fragmentation spectra, they help establish the marker’s
identity.
One of the markers with a m/z of 544.33 (z = +1) showed not only an abundant peak at
m/z 184.07 characteristic of a phosphocholine but, in addition had several product ions in the
fragmentation spectra suggesting it was a sodium (Na) adduct.29 A prominent peak at m/z
485.23, which represented a loss of trimethylamine (TMA, loss of 59.07 mass units) confirmed it
being a Na adduct. This interaction of Na with the lipid makes the negatively charged oxygen
available for nucleophilic reactions, and one common result is the loss of TMA. In addition,
there was a peak at m/z 339.26, generated by loss of sodium-associated phosphocholine (neutral
loss of 205 mass units) from the parent. Further, due to the loss of an acyl-sn-glycerol group
from the fragment ions generated through neutral loss of 59.07, there was a prominent peak at
m/z 146.97. The peak at 104.10 also evidences the PC to be an LPC.30 These findings confirm
the identity of the lipid marker as an LPC with elemental composition of [C26H52NO7P]Na+. This
64
is same molecular formula as the LPC with neutral mass of 521.33 (Figure 2.4). Due to the ina-
bility of mass spectrometry to determine the location of fatty acids and double bonds within fatty
acids, it cannot be confirmed unambiguously that these two molecules represent the same spe-
cies, but it is nonetheless likely. Similarly, an unidentified marker with m/z 566.32 was identified
as a sodiated LPC with elemental composition as [C28H50NO7P]Na+.
The second differentially expressed lipid not identified by reference to LIPID MAPS had
an m/z of 570.34. Fragmentation of this molecule showed an intense peak at 184.07, indicative
of a PC or SM. Exact mass studies determined its elemental composition as C30H52NO7P. Thus,
this lipid molecule is an LPC with a fatty acid 22:5 (docosapentaenoic acid). The general classifi-
cation and constituents of the other lipids are summarized in Table 2.3.
The other differentially expressed lipids (m/z’s: 1159.57, 647.52) did not provide an in-
terpretable MS2 fragment spectra owing to their low initial abundance.
65
Figure 2.3. Summed MS2 spectrum of a differentially expressed sodiated lysophospha-tidylcholine (m/z 544.33, z = +1) (mass accuracy of the precursor peak =51), with its collision product ions and their proposed chemical structures or the proposed compo-nent lost with fragmentation. Mass accuracies of individual product ions having m/z values of 86.09, 104.1, 146.97, 184.06, 258.1, 339.26, 485.23 were 64 ppm, 59 pm, 68 ppm, 42 ppm, 37 ppm, 63 ppm, 53 ppm respectively.
66
Table 2.3. Summary of the chemical classes and components of the lipids differen-tially expressed in chorionic plate versus basal plate of human placenta.
Lipid marker (m/z)
z Neu-tral mass
Lipid class Molecular for-mula
Fatty acid components
520.33 +1 519.37 Glycerophospholipids(LPC)a C26H50NO7P (18:2/0:0)b
542.32 +1 541.31 Glycerophospholipids(LPC) C28H48NO7P (20:5/0:0) 544.33 [M+Na]+ 521.33 Glycerophospholipids(LPC) C26H52NO7P (18:1/0:0) 566.32 [M+Na]+ 543.37 Glycerophospholipids(LPC) C28H50NO7P (20:4/0:0)
or (0:0/20:4)
568.33 +1 567.33 Glycerophospholipids(LPC) C30H50NO7P (22:6/0:0) 570.34 +1 569.33 Glycerophospholipids(LPC) C30H52NO7P (22:5/0:0)
or (0:0/22:5)
673.52 +1 671.52 Phosphosphingolipids C37H73N2O6P (18:2/14:0) 701.55 +1 700.58 Phosphosphingolipids C39H77N2O6P (18:1/16:1)
or (16:1/18:1) or (18:2/16:0)
aLPC: Lysophosphocholine, bThe annotation (e.g. (18:2/0:0)) describes the number of carbon atoms in the fatty acid chain with the number of double bonds. In this case, the fatty acid is composed of 18 carbons with two double bonds at the sn-1 position of the glycerol backbone, while 0:0 represents the absence of a fatty acid chain at the sn-2 position.
67
2.5 Discussion
Tissue proteomics is an increasingly accepted tool for exploring biochemical changes ac-
companying normal physiology and pathophysiology. Global, unbiased proteomic approaches
using multidimensional separations followed by MS are able to identify hundreds to a few thou-
sand proteins but are not quantitative without a single, predetermined target and having a heavy
isotope analogue of the target available which is introduced into the specimen in a defined
amount. Such approaches do not document low abundance proteins or other low abundance spe-
cies. Historically, electrophoretic gel methods, providing some quantitative assessment, have
been coupled to MS for protein identification. More recent comparative and quantitative proteo-
mic approaches include ICAT or iTRAQ where heavy or light isotope labeling of two different
tissue homogenates or cell lysates allow comparisons between those two states or groups.31, 32
Recent advances in top-down proteomics involving multiple separations coupled to high resolu-
tion MS have allowed intact protein sequencing with identification of hundreds of proteins or
proteoforms, including PTMs.5, 6, 33 However, these methods, including top-down methods, still
interrogate only the more abundant, full length proteins. Proteins have obvious importance in cell
or tissue physiology, but small proteins, peptides and lipids are also recognized as mediating or
regulating many intracellular events.34
Peptidomics, including tissue peptidomics, use similar approaches. Both isotopic tagging
(ICAT or iTRAQ) and non-isotopic quantitation techniques have been used.8, 9, 35 Typically, pep-
tides, after their separation from larger proteins, are submitted to liquid chromatography coupled
to tandem MS without trypsin digestion, given they are peptides already. Currently, automated
sequencing, using data-dependent acquisition to select peptides for fragmentation is used in con-
68
junction with a tandem mass spectrometer. This results in the top 3 to 20 (depending on instru-
ment) most abundant peptide species in any initial mass spectrum being submitted to fragmenta-
tion and the fragment profile submitted to search engines to determine amino acid sequence.
Such approaches report dozens of automatically identified sequences,36 but only for those pep-
tides that are high in abundance (the top 3 to top 20 peaks) in the initial MS and only those that
have high enough abundance to provide substantially complete b- and y-fragment ion series in a
single pass and only those in low charge states. De novo sequencing is dramatically more com-
plicated and involves several runs at different fragmentation energies with compilation of data
from those several runs. Typically, these fragments cannot be simply compared to a database.
Yet, this approach is required if you are working with low abundance species. Hence, low abun-
dance species, as monitored in our method, cannot currently be sequenced using automated ap-
proaches and in depth, de novo analysis has been reserved for only those that were quantitatively
different in these studies. However, the approach can be applied to any species of any abundance
if of interest.
Tissue lipidomic approaches are also relatively new and have been reported primarily for
brain tissues.10-12 These typically use organic solvent extraction methods and direct injection alt-
hough liquid chromatography coupled to MS has also been used. Such methods have targeted
those lipids partitioned into the non-polar organic extract and as such they likely sample a some-
what different set of lipids than those interrogated here. Hence, the methods would be comple-
mentary.
With the approach described here a protein depletion step was employed removing highly
abundant species and markedly reducing ion suppression. We interrogated 7000-8000 unique,
novel, low abundance small proteins, peptides and lipids—thousands more than any other current
69
method. Moreover, the great majority of these molecules would not be surveyed by other proteo-
mic, peptidomic or lipidomic approaches due to differences in methodology. Hence, this ap-
proach adds substantially to the array of tissue molecules that can now be studied by MS. The
reproducibility was sufficient 13 to allow for label-free quantitative comparisons. Post transla-
tional modifications and modifications due to reactive oxygen species or alkylating agents can be
observed. Identification or characterization can be applied to any species observed but automated
sequencing procedures would not be adequate for most of the molecules seen in our studies due
to instrument limitations and the low concentrations of species. Having to do intense, multi-
stepped de novo sequencing is not a plus for this approach but one would have to do this regard-
less of method for any of the low abundance, smaller species in a cell or tissue with current MS
instrumentation.
The test of our approach was on placenta. The placenta is interesting clinically but chal-
lenging anatomically. Earlier proteomic studies on human placenta have investigated abundant
proteins, identifying many structural and housekeeping proteins, including actin, other cytoskele-
tal proteins, molecular chaperones, and proteins involved in transport.37 Broadly or narrowly, re-
gional differences within the placenta itself do not appear to have been previously explored using
global proteomic methods but may be very important. Other studies have looked selectively at
trophoblast cells released from placenta.38 The human placenta likely participates in disorders of
pregnancy, including PE,16 but how is incompletely understood.39 There is early evidence that
maternal versus fetal cells may participate in different ways in PE.18 A more in-depth knowledge
of the peptide and lipid repertoire by region in placenta would likely yield additional details of
the pathologic pathways involved, especially if the number of molecules queried were greater.
Our ‘omics’ approach interrogates many thousand additional small proteins, peptides and lipids
70
in a tissue that may participate mechanistically or report changes in tissue function and are meant
to be complementary. Changes in the placental expression by region is likely to be more dra-
matic in PE than in unstimulated or non-pathologic placentae as studied here. Having the ability
to regionally locate these molecular changes may contribute substantially to our understanding of
the placenta’s role in PE.
We undertook to study two regions of placenta: the chorionic plate, composed primarily
of fetal cells, and the basal plate, consisting of a mixture of fetal (trophoblast) and maternal
cells.44 It was recognized that most biomolecules were found in all cells and with overlap of cell
type in these two regions not many differences would be found. Despite this overlap, 16 molecu-
lar species demonstrated statistically significant differences between these two placental regions.
This suggests that the method was robust and reproducible enough to overcome this challenge.
Of these 16, 6 were found to be more abundant in chorionic plate; the remaining 10 were found
to be in higher abundance in the basal plate. These were consistent changes occurring in most of
the 12 placenta characterized independent of each other (Figure 2.5). Fragmentation, characteri-
zation studies were successful in sequencing or chemically classifying 12 markers. Of these, 4
were peptides while the others were lipids.
71
Figure 2.4. Differential expression of peptide with m/z of 718.36 (z = +2). A. This is a line plot of the abundance of the peptide in the basal versus the chorionic plate of each of the 12 replicates demonstrating the consistency of its differential expression. Each colored line represents a single specimen. B. This is a box and whisker plot of the same peptide and its mean abundance in the same two regions of the 12 placentas.
72
One of the peptides was derived from fetal hemoglobin (Hgb γ)40 and more abundant in
the chorionic plate of placenta which is predominantly of fetal origin. Another peptide, with m/z
of 614.38 and a charge state of +3 was identified as a peptide derived from Hgb alpha-1 and was
also more abundant in the chorionic plate, likely in conjunction with the higher Hgb γ present.
Another peptide (m/z = 808.87, z= +2) was a fragment of fibrinopeptide A, phosphory-
lated at serine 3 and more abundant in the chorionic plate. This agrees with previous research.41
Fibrinopeptide A (FbA) was found to be in higher abundance in placenta at term. It has also been
shown that fetal fibrinogen contains twice as much phosphate as adult fibrinogen.41 The biologi-
cal significance of phosphorylation of fetal fibrinogen and of this particular peptide is not
known, but phosphorylation of FbA at Ser-3 has been proposed to produce a more efficient en-
zyme-substrate complex for the highly regulated blood coagulation system as required during
pregnancy.42 The fourth peptide (m/z= 760.4, z= +4) was a fragment of Hgb alpha-2. This pep-
tide was also more abundant in the chorionic plate. The two other peptides could not be se-
quenced due to low concentration and/or higher charge state. These peptide differences then
make physiologic sense.
In addition, regional placental differences in lipids were observed. Most of these lipids
were glycerophosphocholines, PCs, although two were sphingomyelins, SMs and all were more
abundant in the basal plate. A possible explanation for this might be that implantation of the
blastocyst in the uterine wall is accompanied by the endometrial lining being transformed into
decidua becoming maternal cells in the placenta more evident in the basal plate.43 This region
has high lipid content43 and may explain higher concentrations of these in the basal plate.
One limitation of all MS lipidomics is the inability to specify the location of fatty acids as
well as the location of double bonds within unsaturated fatty acid constituents. Due to the very
73
low abundance of our lipid regional markers, it was not possible to use other techniques like
NMR to identify the location of double bonds in the fatty acid chains. Characterization of lipids
and their fatty acids are aided only modestly by the very limited available lipid databases and lip-
idomic-focused software.
In this study we have evaluated a novel ‘omics’ approach to explore lower molecular
weight species present in a complex tissue, placenta, and have identified subtle regional differ-
ences between the chorionic plate, consisting primarily of trophoblast/fetal cells, and the outer
surface of cotyledons or basal plate, consisting of a mixture of fetal and decidual cells. This same
approach then could be used to probe regional changes that may be part of either normal or dis-
ease processes. It is likely possible to determine whether changes arise and progress within a
fixed location or whether they occur broadly across a tissue. Using animal models, this approach
would appear able to document the location and the progression pathway of specific changes in
tissues.
2.6 Conclusion
Collectively, these results demonstrate the feasibility of our low molecular ‘omics’
method to probe thousands of less abundant, lower molecular weight species in tissue and iden-
tify regional differences within the same tissue allowing for study of temporal and spatial
changes as part of both normal and pathologic processes. The method also allowed for relative
quantitative comparisons across many tissues and provided identification of differentially ex-
pressed molecules as part of a single approach. This method should substantially add to other
proteomics or peptidomics methods and allow for a more comprehensive analysis of tissue.
74
2.7 References
1. Chen, J.; Kahne, T.; Rocken, C.; Gatze, T.; Yu, J.; Sung, J. J. Y.; Chen, M.; Hu, P.;
Malfertheiner, P.; Ebert, M. P. A., Proteome Analysis of Gastric Cancer Metastasis by
Two-Dimensional Gel Electrophoresis and Matrix Assisted Laser Desorption/Ionization-
Mass Spectrometry for Identification of Metastasis-Related Proteins. J. Proteome Res.
2004, 3 (5), 1009-1016.
2. Kim, J.; Kim, S. H.; Lee, S. U.; Ha, G. H.; Kang, D. G.; Ha, N. Y.; Ahn, J. S.; Cho, H. Y.;
Kang, S. J.; Lee, Y. J.; Hong, S. C.; Ha, W. S.; Bae, J. M.; Lee, C. W.; Kim, J. W.,
Proteome analysis of human liver tumor tissue by two-dimensional gel electrophoresis and
matrix assisted laser desorption/ionization-mass spectrometry for identification of disease-
related proteins. Electrophoresis 2002, 23 (24), 4142-56.
3. Washburn, M. P.; Wolters, D.; Yates, J. R., Large-scale analysis of the yeast proteome by
multidimensional protein identification technology. Nat. Biotech. 2001, 19 (3), 242-247.
4. Kline, K. G.; Wu, C. C., MudPIT analysis: application to human heart tissue. Methods
Mol. Biol. 2009, 528, 281-93.
5. Catherman, A. D.; Skinner, O. S.; Kelleher, N. L., Top Down proteomics: facts and
perspectives. Biochem. Biophys. Res. Commun. 2014, 445 (4), 683-93.
6. Catherman, A. D.; Durbin, K. R.; Ahlf, D. R.; Early, B. P.; Fellers, R. T.; Tran, J. C.;
Thomas, P. M.; Kelleher, N. L., Large-scale top-down proteomics of the human proteome:
membrane proteins, mitochondria, and senescence. Mol. Cell. Proteomics: MCP 2013, 12
(12), 3465-73.
7. Heroux, M. S.; Chesnik, M. A.; Halligan, B. D.; Al-Gizawiy, M.; Connelly, J. M.; Mueller,
W. M.; Rand, S. D.; Cochran, E. J.; LaViolette, P. S.; Malkin, M. G.; Schmainda, K. M.;
75
Mirza, S. P., Comprehensive characterization of glioblastoma tumor tissues for biomarker
identification using mass spectrometry-based label-free quantitative proteomics. Physiol.
Genomics 2014, 46 (13), 467-81.
8. Wardman, J.; Fricker, L. D., Quantitative peptidomics of mice lacking peptide-processing
enzymes. Methods Mol. Biol. 2011, 768, 307-23.
9. Che, F. Y.; Fricker, L. D., Quantitative peptidomics of mouse pituitary: comparison of
different stable isotopic tags. J. Mass Spectrum 2005, 40 (2), 238-49.
10. Tajima, Y.; Ishikawa, M.; Maekawa, K.; Murayama, M.; Senoo, Y.; Nishimaki-Mogami,
T.; Nakanishi, H.; Ikeda, K.; Arita, M.; Taguchi, R.; Okuno, A.; Mikawa, R.; Niida, S.;
Takikawa, O.; Saito, Y., Lipidomic analysis of brain tissues and plasma in a mouse model
expressing mutated human amyloid precursor protein/tau for Alzheimer's disease. Lipids
Health Dis. 2013, 12, 68.
11. Chan, R. B.; Oliveira, T. G.; Cortes, E. P.; Honig, L. S.; Duff, K. E.; Small, S. A.; Wenk,
M. R.; Shui, G.; Di Paolo, G., Comparative lipidomic analysis of mouse and human brain
with Alzheimer disease. J. Biol. Chem. 2012, 287 (4), 2678-88.
12. Korkes, H. A.; Sass, N.; Moron, A. F.; Câmara, N. O. S.; Bonetti, T.; Cerdeira, A. S.; Da
Silva, I. D. C. G.; De Oliveira, L., Lipidomic Assessment of Plasma and Placenta of
Women with Early-Onset Preeclampsia. PLoS One 2014, 9 (10), e110747.
13. Alvarez, M. T.; Shah, D. J.; Thulin, C. D.; Graves, S. W., Tissue proteomics of the low-
molecular weight proteome using an integrated cLC-ESI-QTOFMS approach. Proteomics
2013, 13 (9), 1400-11.
14. Esplin, M. S.; Merrell, K.; Goldenberg, R.; Lai, Y.; Iams, J. D.; Mercer, B.; Spong, C. Y.;
Miodovnik, M.; Simhan, H. N.; van Dorsten, P.; Dombrowski, M., Proteomic
76
identification of serum peptides predicting subsequent spontaneous preterm birth. Am. J.
Obstet. Gynecol. 2010, 204 (5), 391.e1-391.e8.
15. Merrell, K.; Thulin, C. D.; Esplin, M. S.; Graves, S. W., An integrated serum proteomic
approach capable of monitoring the low molecular weight proteome with sequencing of
intermediate to large peptides. Rapid Commun. Mass Spectrom. 2009, 23 (17), 2685-96.
16. Roberts, D. J.; Post, M. D., The placenta in pre-eclampsia and intrauterine growth
restriction. J. Clin. Pathol. 2008, 61 (12), 1254-1260.
17. Prouillac, C.; Lecoeur, S., The Role of the Placenta in Fetal Exposure to Xenobiotics:
Importance of Membrane Transporters and Human Models for Transfer Studies.
Drug Metab. Dispos. 2010, 38 (10), 1623-1635.
18. Hwang, H. S.; Park, S. H.; Park, Y. W.; Kwon, H. S.; Sohn, I. S., Expression of cellular
prion protein in the placentas of women with normal and preeclamptic pregnancies. Acta
Obstet. Gynecol. Scand. 2010, 89 (9), 1155-61.
19. Damsky, C. H.; Fitzgerald, M. L.; Fisher, S. J., Distribution patterns of extracellular matrix
components and adhesion receptors are intricately modulated during first trimester
cytotrophoblast differentiation along the invasive pathway, in vivo. J. Clin. Invest. 1992,
89 (1), 210-22.
20. Reimel, B. A.; Pan, S.; May, D. H.; Shaffer, S. A.; Goodlett, D. R.; McIntosh, M. W.;
Yerian, L. M.; Bronner, M. P.; Chen, R.; Brentnall, T. A., Proteomics on Fixed Tissue
Specimens - A Review. Curr. Proteomics 2009, 6 (1), 63-69.
21. Robinson, J. M.; Vandre, D. D.; Ackerman, W. E. t., Placental proteomics: a shortcut to
biological insight. Placenta 2009, 30 (9), 13.
77
22. Centlow, M.; Hansson, S. R.; Welinder, C., Differential Proteome Analysis of the
Preeclamptic Placenta Using Optimized Protein Extraction. J. Biomed. Biotechnol. 2010,
2010.
23. Wang, F.; Shi, Z.; Wang, P.; You, W.; Liang, G., Comparative proteome profile of human
placenta from normal and preeclamptic pregnancies. PLoS One 2013, 8 (10), 2013.
24. Dunn, W. B.; Brown, M.; Worton, S. A.; Crocker, I. P.; Broadhurst, D.; Horgan, R.;
Kenny, L. C.; Baker, P. N.; Kell, D. B.; Heazell, A. E., Changes in the metabolic footprint
of placental explant-conditioned culture medium identifies metabolic disturbances related
to hypoxia and pre-eclampsia. Placenta 2009, 30 (11), 974-80.
25. Andersen, H. U.; Fey, S. J.; Larsen, P. M.; Nawrocki, A.; Hejnaes, K. R.; Mandrup-
Poulsen, T.; Nerup, J., Interleukin-1beta induced changes in the protein expression of rat
islets: a computerized database. Electrophoresis 1997, 18 (11), 2091-103.
26. Schröder, S.; Zhang, H.; Yeung, E. S.; Jänsch, L.; Zabel, C.; Wätzig, H., Quantitative Gel
Electrophoresis: Sources of Variation. J. Proteome Res. 2008, 7 (3), 1226-1234.
27. Piehowski, P. D.; Petyuk, V. A.; Orton, D. J.; Xie, F.; Ramirez-Restrepo, M.; Engel, A.;
Lieberman, A. P.; Albin, R. L.; Camp, D. G.; Smith, R. D.; Myers, A. J., Sources of
Technical Variability in Quantitative LC-MS Proteomics: Human Brain Tissue Sample
Analysis. J. Proteome Res. 2013, 12 (5), 2128-2137.
28. Ma, B.; Zhang, K.; Hendrie, C.; Liang, C.; Li, M.; Doherty-Kirby, A.; Lajoie, G., PEAKS:
powerful software for peptide de novo sequencing by tandem mass spectrometry. Rapid
Commun. Mass Spectrum. 2003, 17 (20), 2337-42.
29. Xu, F.; Zou, L.; Lin, Q.; Ong, C. N., Use of liquid chromatography/tandem mass
spectrometry and online databases for identification of phosphocholines and
78
lysophosphatidylcholines in human red blood cells. Rapid Commun. Mass Spectrom. 2009,
23 (19), 3243-54.
30. Al-Saad, K. A.; Siems, W. F.; Hill, H. H.; Zabrouskov, V.; Knowles, N. R., Structural
analysis of phosphatidylcholines by post-source decay matrix-assisted laser
desorption/ionization time-of-flight mass spectrometry. J. Am. Soc. Mass Spectrom. 2003,
14 (4), 373-82.
31. Shi, Z.; Long, W.; Zhao, C.; Guo, X.; Shen, R.; Ding, H., Comparative proteomics analysis
suggests that placental mitochondria are involved in the development of pre-eclampsia.
PLoS One 2013, 8 (5), 2013.
32. DeSouza, L.; Diehl, G.; Rodrigues, M. J.; Guo, J.; Romaschin, A. D.; Colgan, T. J.; Siu, K.
W., Search for cancer markers from endometrial tissues using differentially labeled tags
iTRAQ and cICAT with multidimensional liquid chromatography and tandem mass
spectrometry. J. Proteome Res. 2005, 4 (2), 377-86.
33. Kelleher, N. L., Top-down proteomics. Anal. Chem. 2004, 76 (11), 197A-203A.
34. Gygi, S. P.; Rist, B.; Griffin, T. J.; Eng, J.; Aebersold, R., Proteome analysis of low-
abundance proteins using multidimensional chromatography and isotope-coded affinity
tags. J. Proteome Res. 2002, 1 (1), 47-54.
35. Xu, Z.; Wu, C.; Xie, F.; Slysz, G. W.; Tolic, N.; Monroe, M. E.; Petyuk, V. A.; Payne, S.
H.; Fujimoto, G. M.; Moore, R. J.; Fillmore, T. L.; Schepmoes, A. A.; Levine, D. A.;
Townsend, R. R.; Davies, S. R.; Li, S.; Ellis, M.; Boja, E.; Rivers, R.; Rodriguez, H.;
Rodland, K. D.; Liu, T.; Smith, R. D., Comprehensive Quantitative Analysis of Ovarian
and Breast Cancer Tumor Peptidomes. J. Proteome Res. 2009, 14 (1), 422-433.
79
36. Michalski, A.; Cox, J.; Mann, M., More than 100,000 Detectable Peptide Species Elute in
Single Shotgun Proteomics Runs but the Majority is Inaccessible to Data-Dependent
LC/MS/MS. J. Proteome Res. 10 (4), 1785-1793.
37. Paule, S.; Li, Y.; Nie, G., Cytoskeletal remodelling proteins identified in fetal-maternal
interface in pregnant women and rhesus monkeys. J. Mol. Histol. 2011, 42 (2), 161-6.
38. Orendi, K.; Kivity, V.; Sammar, M.; Grimpel, Y.; Gonen, R.; Meiri, H.; Lubzens, E.;
Huppertz, B., Placental and trophoblastic in vitro models to study preventive and
therapeutic agents for preeclampsia. Placenta 2011, 2011 (32), 023.
39. Mutter, W. P.; Karumanchi, S. A., Molecular mechanisms of preeclampsia. Microvasc.
Res. 2008, 75 (1), 1-8.
40. Bauer, D. E.; Orkin, S. H., Update on fetal hemoglobin gene regulation in
hemoglobinopathies. Curr. Opin. Pediatr. 2011, 23 (1), 1-8.
41. Seydewitz, H. H.; Witt, I., The fraction of high molecular weight (HMW) fibrinogen and
phosphorylated fibrinopeptide A in fetal fibrinogen. Thromb. Res. 1989, 55 (6), 785-90.
42. Maurer, M. C.; Peng, J. L.; An, S. S.; Trosset, J. Y.; Henschen-Edman, A.; Scheraga, H.
A., Structural examination of the influence of phosphorylation on the binding of
fibrinopeptide A to bovine thrombin. Biochemistry 1998, 37 (17), 5888-902.
43. Lu, G.; Zhu, S.; Ke, Y.; Jiang, X.; Zhang, S., Transplantation-potential-related biological
properties of decidua basalis mesenchymal stem cells from maternal human term placenta.
Cell Tissue Res. 2013, 352 (2), 301-12.
44. Wang, Y.; Zhao, S. Vascular Biology of the Placenta. San Rafael (CA): Morgan & Claypool
Life Sciences, 2010
80
3 Global ‘Omics’ Evaluation of Human Placental Responses to Preeclamptic Conditions
This chapter is in large part was reproduced with permission from an accepted research ar-
ticle: Kedia, K.; Smith, S.; Wright, A.H.; Barnes, J.; Tolley, D.H.; Esplin, S.; Graves, S.W.,
Global ‘Omics’ Evaluation of Human Placental Responses to Preeclamptic Conditions. Ameri-
can Journal of Obstetrics & Gynecology
3.1 Abstract
Preeclampsia is a leading cause of maternal death. Its cause is still debated but there is
general agreement that the placenta plays a central role. Perhaps the most commonly proposed
contributors to preeclampsia include placental hypoxia, oxidative stress and an increased pro-in-
flammatory cytokines. How the placenta responds to these abnormalities has been considered but
not as part of a comprehensive analysis of low molecular weight biomolecules and their re-
sponses to these accepted preeclamptic conditions. Using a peptidomic approach, we sought to
identify a set of molecules exhibiting differential expression in consequence of provocative
agents/chemical mediators of preeclampsia applied to healthy human placental tissue. Known PE
conditions were imposed on normal placental tissue from 13 uncomplicated pregnancies and
changes in the low molecular weight biomolecules evaluated. A t-test was used to identify poten-
tial markers for each imposed stress. These markers were then submitted to a LASSO multino-
mial logistic regression model to identify signatures specific to each stressor. Estimates of model
performance on external data were obtained through internal validation. A total of 146 markers
were increased/decreased as a consequence of exposure to proposed mediators of preeclampsia.
81
Of these 75 changed with hypoxia; 23 with hypoxia-reoxygenation/oxidative stress and 48 from
exposure to TNFα. These markers were chemically characterized using tandem mass spectrome-
try. Identification rates were: hypoxia: 34%, hypoxia-reoxygenation 60% and TNFα: 50%.
LASSO modeling specified 16 markers that effectively distinguished all groups, i.e. the 3 abnor-
mal conditions and control. Bootstrap estimates of misclassification rates, multiclass AUC, and
Brier score were 0.108, 0.944, and 0.160, respectively. Using this approach we found previously
unknown molecular changes in response to individual PE conditions that allowed development
biomolecular signatures for exposure to each accepted pathogenic condition.
3.2 Introduction
Preeclampsia (PE) is a disorder of pregnancy characterized by hypertension and pro-
teinuria. Its cause remains unknown. Despite increased understanding of its pathophysiology, PE
incidence has increased in the US over the past decade.1 As many as 75,000 women die world-
wide yearly from PE.2 No established therapeutics exist and efforts to develop such have been
hampered by the incomplete explanation of its cause. Currently, when PE cannot be temporized
by clinical management, the pregnancy is ended. Typically, there is a rapid resolution of the hy-
pertension, proteinuria and other abnormalities. These findings and substantial other research
suggest that the placenta is a necessary for and likely participant in PE.3 Yet, the specific changes
in the placenta that may lead to PE and their antecedent causes are still debated.4
Placentae from women with established PE demonstrate many alterations but it is unclear
which are part of early PE pathogenesis. Among these, some appear to influence placentae very
early to produce PE. These include hypoxia, oxidative stress and increased exposure to pro-in-
flammatory cytokines.5-7 These changes are thought to arise from inadequate remodeling of the
82
maternal vasculature leading to poor placental perfusion.8 Placentae experienceing hypoxia, oxi-
dative stress or inflammatory cytokines, e.g. tumor necrosis factor-α (TNFα) initiate a cascade of
events leading to maternal features of PE.5-7 Knowing what effect these factors might have on
normal placenta could be used to define specific placental changes that should be evident in PE.
Evaluating molecular changes in placenta has frequently involved proteomic analysis of
tissue from women with and without established PE. Common methods primarily identify highly
abundant, high molecular weight proteins (HMW), many being structural proteins and chaper-
ones.9, 10
The aim of this study was to document molecular changes in placenta with exposure to
conditions thought to cause PE. To accomplish this, changes in the abundance of LMW bio-
molecular placental components were determined by a global, tissue ‘omics’ approach for each
condition thought to participate in PE, first demonstrate that changes take place in response to
that pathologic state and second to identify a set or signature of molecular responses characteris-
tic of each. This should allow a PE placenta’s composition to define its exposure to early and
continued abnormalities.
3.3 Materials and methods
3.3.1 Specimen collection
Institutional Review Board approval was obtained from Brigham Young University and
Intermountain Health Care to collect human placentae. No patient personal or medical infor-
mation was recorded in this process and no patient identifiers provided. Thirteen placentas were
collected immediately after elective Caesarian section from uncomplicated term pregnancies. A
full-thickness (~1-inch square) block of placenta was dissected rapidly midway between the cord
83
and placental edge, placed on ice and processed within 30 min of delivery. No placenta came
from a complicated pregnancy or from women with preexisting disease, e.g. hypertension or dia-
betes.
3.3.2 Sample processing
Fetal membranes and decidua were removed. Explants were collected from the intervillous
region representing a point midway between the chorionic and basal plates and ~1 cm thick. Af-
ter initial washing with ice-cold, sterile PBS, the tissue was cut into thin sections (<3mm), kept
on ice to minimize proteolysis. Slices were washed another 8-10 times until nearly all blood was
removed.
3.3.3 Explant culture
Tissue (~300 mg) was cultured in 4 mL Dulbecco's Modified Eagle Medium (DMEM) sup-
plemented with 1% penicillin-streptomycin (Sigma-Aldrich, St. Louis, MO). Normal oxygen ten-
sion in the intervillous region of placenta has been reported to be 6–8%.11, 12 This was considered
normoxic. All explants were maintained in a modular air chamber (Billups-Rothernberg, Del
Mar, CA) filled with 10% oxygen, 5% carbon dioxide, 85% nitrogen for normoxic cultures. The
medium was changed after 24 hr, and control or abnormal conditions were imposed for 48 hr.
Each explant was divided into multiple strips, one incubated under ‘normal’ conditions
and one for each of the PE associated conditions, termed ‘stressors’ as described below:
Hypoxia: Hypoxia was estimated to be 1–2% oxygen for intervillous tissue.12 For hy-
poxic treatment, placental explants were exposed to 2% oxygen, 5% CO2 and 93% N2.
84
Hypoxia-reoxygenation (H/R): Placental explants were cultured under hypoxic condi-
tions for 24 hr and then normoxic conditions for 24 hr. H/R results in release of reactive oxygen
species (ROS).6
Inflammatory cytokines: TNFα was added to the culture medium at a final concentra-
tion of 5.74 nM with incubations of 48 hrs.13, 14
3.3.4 Homogenization of placental explants
Following incubation, individual explants were homogenized as described elsewhere.15
Briefly, 300 mg of placenta were homogenized in the presence of 20 µL of protease inhibitor
cocktail (P9599, Sigma-Aldrich) plus 20 µL of 8.87 mM 1,10-phenanthroline. The homogenate
was resuspended in 3 mL PBS and aliquots treated with two volumes of acetonitrile (1:2,v/v).
This reduced proteins in the specimen, minimizing ion suppression, allowing for ~7,500-8,000
additional low abundance species to be interrogated using mass spectrometry (MS).15
3.3.5 Mass spectrometric analysis
After capillary liquid chromatography15 eluate was introduced into a quadrupole (Q)-time-
of-flight (TOF) tandem mass spectrometer (Q-TOFMS) through an electrospray needle (Agilent
Technologies 6530 Accurate-Mass Q-TOF/LC/MS). Needle voltage was 3800V and samples
were run in the positive ion mode. Mass spectra were collected 8/sec over a mass-to-charge (m/z)
range of 400-3000. Data were acquired using MassHunter Data Acquisition B.06.00 and ana-
lyzed using Qualitative MassHunter B.06.00 software (Agilent Technologies). Other MS param-
eters were: gas temperature: 300oC, drying gas flow rate: 5 L/min, nebulizer pressure: 15 psi,
fragmentor voltage: 175V, skimmer voltage: 65V.
85
Time normalization, as described previously,15 was achieved by defining 2 min windows
across the entire LC chromatogram using 11 time markers to define the middle of each elution
window. These are summarized below.
Elution windows were evaluated separately. Mass spectra for all samples receiving one
treatment were given one color and overlaid with other samples given a second color from the
same placentas maintained under control conditions. In overlays peaks that appeared different
quantitatively (differences of ~1.5x) were evaluated statistically. Peak intensities were deter-
mined using the ‘extract ion count’ (XIC) function of the MS software (MassHunter), compiled
and statistically tested. For any differentially expressed peak, a nearby peak, representing an en-
dogenous species having comparable abundance under stressed and control conditions was also
extracted and peak heights tabulated. These were used as internal references to normalize candi-
date peaks. In the 11 two-minute windows approximately 7,500-8,000 MS features were ob-
served.
3.3.6 Statistical analysis
To reduce the number of potential peaks from which to identify potential biomarkers and
to build a classification model, Student’s t-test (two-tailed, paired) was performed on each peak
in the data set to determine whether the abundance of an extracted peak was different between
the control and a given condition. A p-value <0.05 was considered indicative of a potential bi-
omarker. Any peak with a p-value <0.05 in any comparison (i.e. controls vs. hypoxia, controls
vs. hypoxia-reoxygenation, controls vs. TNFα) was retained as a possible biomarker for the crea-
tion of a classification model.
86
To build the classification model, LASSO multinomial logistic regression16 was selected
due to its high internal validation performance relative to other methods and its inherent ability
to perform peak selection. The final model identified 16 peaks that were effectively able to dis-
tinguish the 4 classes of treated placentas (control, hypoxia, hypoxia-reoxygenation, and inflam-
matory cytokines). Table 3.2A summarizes the selected peaks and model coefficients. On the
training data, the model had a correct classification rate of 100%. Note the distinct regions in
Figure 3.2 that each group occupies, another indication of model’s ability to discriminate be-
tween the four treatments (https://CRAN.R-project.org/package=plot3D). Since no external data
was available for an external validation of the classification model, the performance was as-
sessed through internal validation. To assess overall model performance, we applied the boot-
strap method17 to obtain estimates of the multiclass AUC18 and the Brier score19. We used the
.632+ bootstrap method to estimate the misclassification rate.20 As seen in Table 3.2B, the model
is expected to have very good performance. We note that since only the peaks in the list of po-
tential biomarkers were considered in building the final model, the estimates from internal vali-
dation are expected to be optimistic. All analyses were performed using R (R Core Team 2015,
https://www.R-project.org/).
3.3.7 Peptide identification
Peptide candidates were submitted to tandem MS fragmentation studies to sequence them
using N2 and sometimes Ar (for poorly fragmenting species) as the collision gas.15 The acquisi-
tion rate for MS/MS experiments was 1 scan/sec and the isolation window was 1.3 m/z for low-
charge and 4 m/z for high-charge state peptides.
87
Due to low initial marker abundance, fragmentation data did not provide complete b- and
y-series and de novo sequencing was required.
3.3.8 Lipid identification
Lipids were fragmented similarly using tandem MS. Databases LIPID Maps
(http://www.lipidmaps.org/), Metlin (https://metlin.scripps.edu/), and Human Metabolome Data-
base (HMDB) (http://www.hmdb.ca/) were used to assign lipid structures based on fragment pat-
terns. For markers without matches in databases, elemental composition was predicted by exact
mass studies using commercial standards having m/z’s 121.0509 and 922.0098 (Agilent Tech-
nologies).
3.4 Results
3.4.1 Chromatographic time alignment
The 11 time normalization peaks are summarized by their m/z, charge state and approxi-
mate elution time as follows: 695.09 (z=+4) 17 min, 827.78 (z=+6) 19 min, 474.2 (z=+1) 21 min,
672.36 (z=+3) 23 min, 686.46 (z=+1) 25 min, 1009.05 (z=+1) 27 min, 616.16 (z=+1) 32 min,
526.28 (z=+1) 34 min, 524.36 (z=+1) 37 min, 650.43 (z=+1) 39 min, 675.53 (z=+1) 41 min.
3.4.2 Low abundance, LMW weight species altered by PE conditions
Most molecules did not change significantly with treatment. However, several biomole-
cules changed in response to each treatment. A few changed with more than one. For example,
88
peptide m/z 827.78 (z=+6) decreased with hypoxia and hypoxia-reoxygenation but not TNF-α
(Figure 3.1.A-B)
89
Figure 3.1. MS overlay and box-and-whisker plot for peptide m/z 827.78 (=+6) A. Over-lay of 26 mass spectra in the region containing peptide m/z 827.78 (z=+6) with its iso-tope envelope. Blue, placental explants under control conditions. Red, placental ex-plants under hypoxic conditions. Species less abundant in the hypoxic placental ex-plants. B. Box-and-whisker plot of peptide m/z 827.78 (z=+6)
In response to hypoxia significant changes were observed for 75 markers (Table 3.1A).
H/R elicited significant changes in 23 molecules (Table 3.1B). TNFα caused significant changes
in 48 molecules (Table 3.1C). To better understand the biology underlying these changes, tan-
dem MS studies were conducted to chemically classify or identify candidates.
90
Table 3.1. List of differentially expressed biomolecules for all the three ‘stressed’ conditions. A. Hypoxia, B. Hypoxia-reoxygentation, C. TNFα
A. Differentially expressed biomolecules in placental explants cultured under hypoxia M/Z Z M+H P-value Normalized p-value Hypoxia Control
456.17 1 456.17 8.50E-02 3.43E-02 ↓ ↑ 517.03 1 517.03 3.56E-02 3.57E-03 ↑ ↓ 639.33 1 639.33 1.04E-01 4.57E-02 ↓ ↑ 723.22 1 723.22 7.26E-02 1.96E-02 ↓ ↑ 889.57 1 889.57 7.94E-02 2.92E-02 ↓ ↑ 981.43 1 981.43 1.01E-01 3.08E-02 ↓ ↑
489.61* 3 1466.83 3.67E-02 1.83E-02 ↓ ↑ 643.59* 4 2571.36 5.93E-02 2.87E-02 ↓ ↑
520.3 5 2597.5 8.07E-03 3.82E-03 ↓ ↑ 574.3 5 2867.5 4.29E-02 2.16E-02 ↓ ↑
91
830.4* 6 4977.4 1.44E-03 3.87E-04 ↓ ↑ 478.17 1 478.17 1.29E-03 3.00E-03 ↓ ↑ 495.05 1 495.05 8.78E-02 1.13E-01 ↑ ↓ 512.18 1 512.18 1.03E-02 7.77E-03 ↓ ↑ 545.12 1 545.12 9.85E-04 1.41E-03 ↓ ↑ 570.55 4 2279.2 4.98E-03 5.83E-03 ↓ ↑ 736.08 6 4411.48 1.42E-02 2.79E-02 ↓ ↑ 676.49 7 4729.43 8.69E-04 4.54E-03 ↓ ↑ 694.64 7 4856.48 2.18E-04 7.38E-04 ↓ ↑
823.26* 6 4934.56 8.31E-03 3.34E-02 ↓ ↑ 827.78* 6 4961.68 5.28E-06 1.08E-04 ↓ ↑
997.5 5 4983.5 2.04E-03 1.11E-02 ↓ ↑ 1001.9032 5 5005.516 1.06E-03 1.57E-02 ↓ ↑ 1257.62 4 5027.48 5.87E-03 2.93E-02 ↓ ↑
1012.89* 5 5060.45 3.28E-04 5.64E-03 ↓ ↑ 1271.37 4 5082.48 7.46E-03 4.28E-02 ↓ ↑ 1063.73 5 5314.65 7.80E-02 6.09E-03 ↓ ↑ 1334.9 4 5336.6 4.55E-02 9.65E-03 ↓ ↑ 500.14 10 4992.4 1.25E-02 6.27E-03 ↓ ↑ 405.18 1 405.18 5.74E-02 3.54E-02 ↑ ↓ 425.13 1 425.13 1.66E-03 6.65E-04 ↑ ↓ 431.05 1 431.05 7.45E-02 8.62E-02 ↓ ↑ 447.11 1 447.11 3.04E-04 4.13E-06 ↑ ↓ 469.09 1 469.09 1.17E-02 2.51E-02 ↑ ↓ 504.23 1 504.23 5.51E-03 1.49E-02 ↓ ↑ 736.39 6 4413.34 1.43E-04 1.13E-02 ↓ ↑ 427.18 1 427.18 1.55E-02 4.73E-03 ↑ ↓
458.25* 1 458.25 2.45E-03 7.80E-04 ↓ ↑ 563.29 1 563.29 8.32E-03 3.61E-03 ↓ ↑ 400.91 3 1200.73 6.33E-05 2.09E-05 ↓ ↑ 401.57 3 1202.71 3.52E-04 9.91E-05 ↓ ↑
744.49* 11 8179.39 7.71E-05 6.96E-05 ↓ ↑ 711.87 14 9953.18 1.47E-03 1.81E-03 ↓ ↑ 415.19 1 415.19 7.90E-02 1.23E-02 ↑ ↓ 519.27 1 519.27 1.24E-02 1.14E-03 ↓ ↑ 650.97 13 8450.61 5.13E-04 3.29E-04 ↓ ↑ 462.59 3 1385.77 1.01E-02 1.43E-02 ↓ ↑ 868.07 6 5203.42 3.14E-04 6.70E-05 ↓ ↑
753.52* 7 5269.69 4.53E-02 6.88E-02 ↓ ↑ 893.79 6 5357.74 2.25E-02 7.00E-02 ↓ ↑
1038.8* 8 8306.22 4.43E-02 1.14E-01 ↓ ↑
92
429.19 1 429.19 3.93E-02 8.1E-02 ↑ ↓ 435.19 1 435.19 8.62E-03 2.0E-02 ↑ ↓ 413.19 1 413.19 6.07E-02 0.029 ↑ ↓ 470.74 1 470.74 5.91E-02 4.28E-02 ↓ ↑
417.33* 1 417.33 3.16E-03 2.79E-04 ↓ ↑ 424.33* 1 424.33 1.27E-03 3.25E-05 ↑ ↓ 431.3* 1 431.3 1.33E-02 1.26E-03 ↓ ↑
722.41* 1 722.41 5.63E-02 1.33E-02 ↓ ↑ 536.32* 1 536.32 2.26E-02 4.45E-02 ↓ ↑ 616.35 1 616.35 8.87E-02 1.07E-01 ↓ ↑
622.39* 1 622.39 2.19E-02 6.63E-02 ↓ ↑ 652.41* 1 652.41 4.66E-02 5.04E-02 ↓ ↑ 666.42* 1 666.42 2.67E-02 2.97E-02 ↓ ↑ 688.4* 1 688.4 5.08E-02 4.22E-02 ↓ ↑
864.54* 1 864.54 6.34E-02 6.22E-02 ↓ ↑ 594.36* 1 594.36 1.34E-02 3.69E-02 ↓ ↑ 414.76 1 414.76 3.84E-02 2.56E-01 ↓ ↑ 423.75 1 423.75 3.15E-02 1.45E-01 ↓ ↑
452.37* 1 452.37 9.27E-03 9.97E-05 ↑ ↓ 454.38* 1 454.38 8.91E-04 3.46E-06 ↑ ↓ 510.38* 1 510.38 2.99E-02 6.35E-02 ↓ ↑ 480.39* 1 480.39 1.01E-02 7.21E-06 ↑ ↓ 484.43* 1 484.43 2.87E-05 1.05E-06 ↑ ↓ 649.49 1 649.49 3.84E-02 7.57E-03 ↑ ↓
*Identfied and characterized biomolecules described in Table 3 with their identitities.
B. Differentially expressed biomolecules in placental explants cultured under hypoxia-reox-ygentation
M/Z Z M+H P-value Normalized p-value H/R Control
538.53* 4 2151.12 6.38E-04 2.41E-03 ↓ ↑ 823.26* 6 4934.56 9.50E-04 2.46E-05 ↓ ↑ 621.07* 8 4961.56 2.40E-06 4.24E-06 ↓ ↑ 623.19 8 4978.52 1.31E-02 8.23E-03 ↓ ↑ 831.58 6 4984.48 2.31E-05 9.64E-06 ↓ ↑
1012.89* 5 5060.45 3.79E-04 4.36E-04 ↓ ↑ 891.11* 6 5341.66 4.39E-03 9.50E-03 ↓ ↑ 425.13 1 425.13 1.53E-02 2.15E-02 ↑ ↓
93
447.11 1 447.11 9.69E-04 2.70E-03 ↑ ↓ 435.27 3 1303.81 8.30E-02 5.37E-02 ↑ ↓ 458.25* 1 458.25 9.44E-03 8.0E-03 ↓ ↑ 659.51 12 7903.12 3.23E-04 4.5E-03 ↓ ↑ 555.78 1 555.78 1.58E-03 3.14E-02 ↓ ↑ 558.28 1 558.28 5.96E-02 2.66E-02 ↓ ↑ 506.83 2 1012.66 7.78E-02 7.60E-02 ↓ ↑ 462.59 3 1385.77 3.43E-02 8.91E-03 ↓ ↑ 451.74 1 451.74 3.85E-02 1.46E-02 ↓ ↑ 431.31 1 431.31 4.27E-04 3.39E-04 ↓ ↑ 542.31* 1 542.31 7.87E-02 2.86E-02 ↑ ↓ 590.32* 1 590.32 7.81E-03 1.01E-02 ↑ ↓ 566.37 1 566.37 8.16E-02 5.33E-03 ↓ ↑ 426.35* 1 426.35 2.47E-02 3.02E-03 ↑ ↓ 401.26* 1 401.26 4.30E-03 2.47E-03 ↑ ↓
*Identfied and characterized biomolecules described in Table 3 with their identitities.
C. Differentially expressed biomolecules in placental explants cultured in TNFα environ-ment
M/Z Z M+H P-value Normalized p-value TNFα Control
461.06 1 461.06 4.46E-02 1.08E-01 ↑ ↓ 501.19 1 501.19 5.90E-02 1.13E-01 ↓ ↑ 517.03 1 517.03 5.03E-02 2.96E-02 ↑ ↓ 525.28 1 525.28 1.53E-02 5.37E-01 ↓ ↑ 661.29 1 661.29 9.88E-02 2.25E-01 ↓ ↑ 705.24 1 705.24 5.47E-02 7.35E-01 ↓ ↑ 830.4 6 4977.4 1.83E-02 1.16E-01 ↓ ↑ 619.27 1 619.27 5.08E-02 6.49E-02 ↓ ↑ 409.29 1 409.29 1.10E-02 6.81E-03 ↓ ↑ 427.3 1 427.3 1.22E-02 6.38E-03 ↓ ↑ 629.3 1 629.3 9.55E-02 8.91E-02 ↓ ↑ 673.25 1 673.25 7.73E-02 9.43E-02 ↓ ↑
94
801.39 1 801.39 3.33E-02 4.92E-02 ↓ ↑ 853.59 1 853.59 8.76E-03 7.17E-03 ↓ ↑ 832.59 6 4990.54 3.72E-02 4.17E-03 ↑ ↓ 505.25 1 505.25 1.06E-02 8.58E-03 ↓ ↑ 549.27 1 549.27 3.67E-02 9.29E-03 ↓ ↑ 584.94 1 584.94 3.75E-03 3.79E-04 ↑ ↓ 586.97 1 586.97 5.79E-03 2.56E-04 ↑ ↓ 666.14* 5 3326.7 4.77E-02 1.32E-02 ↑ ↓ 458.25* 1 458.25 1.09E-03 4.29E-04 ↑ ↓ 479.25 1 479.25 1.17E-02 1.16E-01 ↑ ↓ 773.15 1 773.15 4.90E-03 1.56E-02 ↓ ↑ 460.27 2 919.54 6.88E-02 7.80E-02 ↑ ↓ 409.16 1 409.16 2.12E-02 3.83E-02 ↓ ↑ 467.12 1 467.12 3.20E-02 3.78E-02 ↓ ↑ 476.24 1 476.24 1.32E-02 2.00E-03 ↑ ↓ 590.22 1 590.22 3.29E-03 5.19E-03 ↓ ↑ 454.28* 1 454.28 6.26E-03 1.10E-01 ↑ ↓ 460.24 1 460.24 1.51E-02 1.22E-01 ↑ ↓ 474.25 1 474.25 1.78E-02 7.42E-02 ↑ ↓ 476.27* 1 476.27 4.63E-02 2.00E-01 ↑ ↓ 494.31* 1 494.31 6.17E-03 2.37E-02 ↑ ↓ 500.27 1 500.27 5.26E-03 2.59E-02 ↑ ↓ 540.25 1 540.25 7.44E-03 4.51E-02 ↑ ↓ 518.31* 1 518.31 9.69E-02 2.37E-01 ↑ ↓ 478.29* 1 478.29 5.05E-03 1.18E-02 ↑ ↓ 480.38 1 480.38 1.62E-02 1.58E-02 ↑ ↓ 520.33* 1 520.33 6.58E-03 1.17E-02 ↑ ↓ 522.33* 1 522.33 1.78E-02 1.43E-02 ↑ ↓ 542.31* 1 542.31 7.53E-03 9.05E-03 ↑ ↓ 544.33* 1 544.33 2.36E-02 2.10E-02 ↑ ↓ 566.32* 1 566.32 6.80E-03 9.23E-03 ↑ ↓ 568.33* 1 568.33 3.12E-02 2.48E-02 ↑ ↓ 997.61* 1 997.81 3.16E-02 1.43E-02 ↑ ↓ 1045.65* 1 1045.65 1.86E-02 1.86E-02 ↑ ↓ 1504.85* 1 1504.85 4.58E-02 4.05E-02 ↑ ↓ 1021.61 1 1021.61 8.53E-02 6.09E-02 ↑ ↓ 502.28* 1 502.28 2.99E-02 5.17E-02 ↑ ↓ 425.28 1 425.28 2.45E-02 2.47E-02 ↑ ↓
*Identfied and characterized biomolecules described in Table 3 with their identitities.
95
3.4.3 Sequenced peptides
Most stress-responsive peptides had charge states >+3 and presented at low abundance
and/or with post-translational modifications (PTMs). Sequences were evaluated by on-line data-
base comparison (Mascot (http://www.matrixscience.com/search_form_select.html, Spectrum
Mills) but ultimately required de novo sequencing.
As an example, peptide m/z 744.49, z=+11 was a ubiquitin peptide based on two partial
de novo amino acid sequences submitted to a BLAST search. Several fragments could be as-
signed to the y-ion series, but none matched the predicted b-series. There was an offset of 132.04
mass units suggesting a pentose modification. After adding this modification to the b1 ion, gluta-
mine, many m/z’s observed in the fragment spectrum then matched predicted ions in the b-series.
96
Glutamine has not been reported to have a pentose modification, but the b1-b4 ions had the mod-
ification present suggesting the modification.
3.4.4 Lipid classification
All lipids were in the +1 charge state. Certain lipids were readily classified by fragments,
e.g. Phosphatidylcholines (PCs) and phosphoethanolamines (PEts) showed signature peaks at
m/z 184.07 or a neutral loss of 141.03 respectively. Additionally acylcarnitines had signature
peaks at m/z 60.04 and 85.02.
Other lipid molecules, e.g. vitamin D21 and 10,11-Dihydro-12R-hydroxy-leukotriene E422
were assigned based on accurate masses and characteristic water losses. These assignments were
confirmed by prior literature.21, 22
Several PCs appeared oxidized, probably at double bonds. Oxidized PCs typically
showed one or two water loss peaks (-18 and -36 mass units) from the precursor ion upon frag-
mentation. Some were confirmed using LIPID MAPS while others were confirmed using exact
mass studies.
Two markers appeared to be dimers of PC and PE and one a trimer of PE based on frag-
mentation patterns. For a dimer of PC and PE, a very abundant peak at the PC’s m/z was present
accompanied by an abundant 184.07 peak. A smaller peak representing a PE was also present. In
all three cases it was PE m/z 502.28 with neutral loss of 141.01 manifest by a peak at m/z 361.27
for all 3 markers. These assignments were confirmed by exact mass studies.
97
3.4.5 Condition specific signatures
A final model identified 16 peaks that effectively segregated the 4 placental treatments
(control, hypoxia, hypoxia-reoxygenation, TNFα) (Table 3.2A). On the training data, the model
had a correct classification rate of 100% (Figure 3.2). The X, Y, and Z axes are not orthogonal,
but represent linear combinations of the model coefficients of Controls, Hypoxia, and Hypoxia-
Reoxygenation treatments, respectively. Bootstrap estimates of misclassification rates, multiclass
AUC, and Brier score were 0.108, 0.944, and 0.160, respectively, suggesting very good perfor-
mance (Table 3.2B). However, because not all molecules were used for analysis (only significant
ones), the estimates may be optimistic.
98
Figure 3.2. Scatterplot of Group Membership Based on the Statistical Model. The num-bers, 1, 2, 3, and 4, correspond to the Controls, Hypoxia, Inflammatory Cytokines, and Hypoxia-Reoxygenation classes, respectively. The x, y, and z axes, which are not con-strained to be orthogonal, represent the predicted probabilities from the lasso model for the Controls, Hypoxia, and Hypoxia-Reoxygenation treatment classes, respectively.
99
Table 3.2. Condition specific changes in biomolecules
3.2A. Non-zero Peaks (M+H value) and Coefficients from Lasso Multinomial Logistic Regression Model
Class M+H Peak Values and Coefficients Control Peak
Coefficient Intercept -2.382
431.3 25.278
452.75 -3.309
484.43 -5.226
1218.73 0.143
2151.12 296.642
2571.36 0.00013
4934.56 1.129
Hypoxia Peak Coefficient
Intercept 0.449
425.13 0.0051
447.11 0.006
452.75 12.911
478.17 -35.301
484.43 6.873
512.18 -0.417
2151.12 -122.656
2571.36 0.000048
TNF-α Peak Coefficient
Intercept 2.561
2571.36 -0.0001
4934.56 -0.633
4977.46 0.766
5082.48 -29.168
Hypoxia-Reoxygen-ation
Peak Coefficient
Intercept -0.627
2571.36 -0.00005
3326.7 0.0016
5341.66 2.234
7903.12 -0.0991
3.2B. Model Performance Based on Internal Validation
Misclassification Rate AUC Brier
0.108 0.944 0.160
100
3.5 Discussion
The cause of PE remains elusive. Current research implicates the placenta in its patho-
genesis. The best documented PE abnormalities experienced by the placenta are hypoxia or H/R
or exposure to pro-inflammatory cytokines.5-7
This study tested two hypotheses. First, it was proposed that normal placental explants
exposed to PE abnormalities would show significant changes in LMW cellular components, and
second, that each abnormality would produce a specific pattern of changes. If these hypotheses
proved true, then it might be possible to find these same patterns in actual PE placenta and thus
provide evidence for a specific abnormality being part of that pregnancy. When hypoxia, H/R or
TNFα was presented to human placental explants, several LMW peptides and lipids demon-
strated significant changes for each. Some changed in response to all three abnormalities, but
most changed in response to only one, suggesting more stressor-specific changes.
Others have studied the placenta as related to PE using cell culture.23 Those studies em-
ployed choriocarcinoma cells and may have limitations due to the use of cancer cells. Explanted
tissue, as used here, is also imperfect, but begins with normal placental tissue and maintains the
native tissue architecture. This may preserve important processes including cell-to-cell commu-
nication and may be closer to the in-vivo environment. As a practical consideration, tissue pro-
vides more material to work with, allowing for better MS coverage of low abundance species.
There are many ‘omics’ methods. Some have been applied to PE placenta, primarily 2-
DGE with excision of differentially expressed proteins, trypsin digestion followed by MS.24 This
identifies full length proteins, but surveys only dozens to a few hundred more abundant species,
101
omitting less abundant proteins and all peptides and lipids. More comprehensive, global prote-
omics methods exist and increase the number of proteins surveyed. They have been applied to
trophoblast cells or secreted products of trophoblast cells after labeling.25 They still measure the
more abundant proteins and omit low abundance proteins, peptides and lipids. A few studies tar-
geting specific proteins or metabolites using immunoassays or MS have been reported.26, 27
Our study, in contrast, focused on less abundant, LMW biomolecules using a global tis-
sue ‘omics’ approach. There is a greater recognition that smaller, less abundant molecules con-
tribute to cell physiology.28 Obviously, interrogation of them will provide a more complete pic-
ture of tissue biology. A second novel aspect of this work is in defining sets of molecular
changes that indicate placental exposure to specific preeclamptic stress conditions. This may pro-
vide biological insights into placental responses but can provide a molecular signature unique to
each abnormality. These signatures may be usefully compared with similarly obtained ‘omic’
profiles found in PE placenta to potentially define the pathologic background of that placenta.
As predicted, several peptides and lipid molecules changed in response to each PE abnor-
mality
The responses to a specific PE condition were consistent across placentae and often more
than one molecule having similar structure or biologic function changed in concert increasing the
likelihood that changes were real. Additionally, several of the significant biomarkers in our study
have been linked with PE previously in the literature. Again, this consistency further suggests
their involvement in the pathophysiology of PE.
The responding markers are organized by function in Table 3.3. This categorization is
similar to a systems biology approach in which several related biological pathways appear to
change in a uniform direction in response to a particular stress.
102
Table 3.3. Summary of categorized markers with their identifications and biological significance
Cytoskeletal elements
m/z Parent pro-tein
Stress Higher in
Sequencec Biological pathways
489.61 (+3)
Vimentin Hypa Ctrb slplvdthskrtl Vimentin helps regulate and stabilize cell adhesion. Vimentin fragments in PE re-main unstudied. Keratin plays a role in im-plantation and placentation.29 Under bio-logical stress, degradation of keratin oc-curs30 with fragments entering the maternal circulation.31
643.59 (+4)
Keratin-8 acetylated (position of modification -unknown)
Hyp Ctr sirvtqksykvstsgprafssrs
830.4 (+6)
Acetylated at N termini-Thymosin B-4(oxi-dized at me-thionine-6)
Hyp, H/R, TNFα
Ctr (Ac)dsdkpdm(Ox)eaeiekfdk-sklkktetqeknplpsketieqekqages
Thymosin appears cytoprotective and angi-ogenic in pregnancy,32 but is unstudied in PE. Our findings differ from one study that found increased thymosin b4 in a hypoxic mouse model.33 Cells in that study were exposed to 1 hr Hyp. Differences in time, tissue, species might account for different findings. The oxidized methionine form of thymosin b-4 may have anti-inflammatory properties.34
823.26 (+6)
Thymosin Beta-10 (Acetylated at N termini)
Hyp & H/R
Ctr (Ac)ad-kpdmgeiasfdkaklkktetqekntlpt-ketieqekrseis
827.78 (+6)
Thymosin Beta-4 (Acetylated at N termini)
Hyp & H/R
Ctr (Ac)sdkpdmaeiekfdk-sklkktetqeknplpsketieqekqages
1012.89 (+5)
Thymosin Beta-4 (Acetylated at N termini; phosphory-lated)
Hyp & H/R
Ctr Sdkpdmaeiekfdksklkktetqeknplpsketieqekqages
891.11 (+6)
thymosin beta 4+380.16
H/R Ctr
103
538.53 (+4)
Fragment of Thymosin B-4 (acety-lated at N)
H/R Ctr sdkpdmaeiekfdk-sklk
Hemoglobin
m/z ID Stressor Higher in
Sequence Biological pathways
753.52 (+7)
Hgbf alpha 1 globin chain
Hyp Ctr
vlspadktn vkaaw-gkvga hageygaeal ermflsfptt ktyfphfdl
Unknown
1038.89 (+8)
Hgb alpha 1 globin chain
Hyp Ctr vlspadktn vkaaw-gkvga hageygaeal ermflsfptt ktyfphfdls hgsaqvkghgkkvadaltna vahvddmpn
666.14 (+5)
Hgb alpha 1 globin chain
TNFα TNFα vlspadktnvkaaw-gkvgahageygaealerm
Ubiquitin
m/z ID Stressor Higher in
Sequence Biological pathways
744.49 (+11)
Ubiquitin (Pentose modification at b1)
Hyp Ctr
qifvktltgk-titlevepsdtien-vkakiqdkegippdqqrli-fagkqledgrtlsdyn iqkestlhlvlr
Hypoxia-inducible factor 1a (HIF-1a) levels are normally low due to its degradation by the ubiquitin system.35 Under hy-poxic conditions, HIF-1a has a longer half-life and increased levels due to decreased ubiquiti-nation.35
Inflammatory Factors
m/z ID Stressor Higher in
Chemical composition [M+H]
Biological pathways
458.25 (+1)
10,11-Dihydro-12R-hydroxy-leu-kotriene E4
Hyp, H/R, TNFα
Ctr TNFα
C23H40NO6S This compound is an oxidized metabolite of LTE4, a mediator of inflammation,36 and its reduction with Hyp and H/R may
104
(LTE4)(Epsilon-LTB(,3))
be due to limited oxygen, while the over-expression with exposure to TNFα may cause deleterious effects.37
Metabolic mediators
m/z ID Stressor Higher in
Chemical composition [M+H]
Biological pathways
417.33 (+1)
Dihydroxyvitamin D3
Hyp Ctr
C27H45O3 Active dihydroxyvitamin D3 helps with decidualization and implantation in preg-nancy.38 Decreased synthesis of D3 pro-duces more Th1 type cytokines adversely affecting the pregnancy.39 PE is associated with low levels of D3.39
Fatty acid metabolism
m/z ID Stressor Higher in
Chemical composi-tion [M+H]
Biological pathways
424.33 (+1)
Fatty acyl car-nitines 9,12-Hexadecadi-enylcarnitine
Hyp Hyp C25H46NO4 Carnitine shuttles fatty acids into the mitochondrial matrix for β-oxidation. Oxygen is required for active electron. Hyp de-creases electron transport in-creasing NADH and FADH2.40 This results acyl-carnitines accu-mulation.41 Studies report in-creased acyl-carnitines in PE.42
452.37 (+1)
Fatty acyl car-nitines (11Z,14Z)-eicosa-dienoylcarnitine
Hyp Hyp C27H50NO4
454.38 (+1)
Fatty acyl car-nitines (11Z)-eicosene-oylcarnitine
Hyp Hyp C27H52NO4
480.39 (+1)
Fatty acyl car-nitines (13Z,16Z)-docosa-dienoylcarnitine
Hyp Hyp C29H54NO4
484.43 (+1)
Fatty acyl car-nitines
Hyp Hyp C29H58NO4
105
O-be-henoylcarnitine
426.35 (+1)
Fatty acyl car-nitines O-oleoylcarnitine
H/R H/R C25H48NO4
Phosphatidylcholines
m/z ID Stressor Higher in
Chemical composi-tion [M+H]
Biological pathways
722.41 (+1)
Either an oxidized PC or normal
Hyp Ctr C38H61NO10P or C42H61NO7P
Hyp may reduce PC synthesis due to less ATP.43, 44 Our studies found increases in several lyso-PCs and lyso-PEs in placenta presented with H/R and TNFα. This is consistent with previous findings of increased phospho-lipase A2 activity in both these conditions with increased lyso-PC and lyso-PE.45, 46 Others re-port higher phospholipase A2 ac-tivity in PE.47
536.32 (+1)
PC (16:1/2:0)g Hyp Ctr C26H51NO8P
594.36 (+1)
Oxidized PC(16:0/5:0(CHO))
Hyp Ctr C29H57NO9P
622.39 (+1)
Oxidized PC (Not on any lipid data-bases)
Hyp Ctr C31H61NO9P
652.41 (+1)
Oxidized PC (Not on any lipid data-bases)
Hyp Ctr C32H63NO10P
666.42 (+1)
Oxidized PC (16:0/9:0(COOH))
Hyp Ctr C33H65NO10P
688.4 (+1)
Sodiated Oxidized PC [M+Na] for m/z 666.42
Hyp Ctr C33H64NO10NaP
864.54 (+1)
Oxidized PC (Not on any lipid data-bases)
Hyp Ctr C48H83NO10P
510.38 (+1)
LPCh (O-18:0/0:0) Hyp Ctr C26H57NO6P
566.31 (+1)
M+Na of 544.33 LPC(20:4/0:0)
H/R H/R C28H50NO7NaP
106
590.32 (+1)
M+Na of 568.33 LPC(22:6/0:0)
H/R H/R C30H50NO7NaP
542.31 (+1)
M+Na of 520.33 LPC(18:2/0:0)
H/R H/R C26H50NO7NaP
494.31 (+1)
LPC (16:1/0:0) TNFα TNFα C24H49NO7P
518.31 (+1)
M+Na of 496.33 LPC (16:0/0:0)
TNFα
TNFα
C24H50NO7NaP
520.33 (+1)
LPC (18:2/0:0) TNFα TNFα C26H51NO7P
522.33 (+1)
LPC (18:1/0:0) TNFα TNFα C26H53NO7P
542.31 (+1)
M+Na of 520.33 LPC (18:2/0:0)
TNFα TNFα C26H50NO7NaP
544.33 (+1)
LPC (20:4/0:0) TNFα TNFα C28H51NO7P
566.32 (+1)
M+Na of 544.33 LPC (20:4/0:0)
TNFα TNFα C28H50NO7NaP
568.33 (+1)
LPC (22:6/0:0) TNFα TNFα C30H51NO7P
Phosphotadylethanolamine
m/z ID Stressor Higher in
Chemical composi-tion [M+H]
454.28 (+1)
LPEi (16:0/0:0) TNFα TNFα C21H45NO7P
476.27 (+1)
LPE (18:3/0:0) TNFα TNFα C23H43NO7P
478.28 (+1)
LPE (18:2/0:0) TNFα TNFα C23H45NO7P
500.27 (+1)
LPE (20:5/0:0) TNFα TNFα C25H43NO7P
107
474.25 (+1)
M+Na of 452.26 PE (16:1/0:0)
TNFα TNFα C21H42NO7NaP
478.29 (+1)
LPE (18:2/0:0) TNFα TNFα C23H45NO7P
502.28 (+1)
LPE (20:4/0:0) TNFα TNFα C25H45NO7P
480.3 (+1)
LPE (18:1/0:0) TNFα TNFα C23H47NO7P
558.35 (+1)
PEt (Not on any lipid databases)
H/R H/R C29H53NO7P
Element composition of unknown markers
m/z ID Stressor Higher in
Chemical composi-tion [M+H]
Biological pathways
431.31 (+1)
Unknown Hyp Ctr C27H43O4 Unknown
401.26 (+1)
H/R H/R C25H37O4
409.16 (+1)
TNFα Ctr C24H25O6
Dimers
m/z ID Stressor Higher in Biological pathways
1045.61 (+1)
Dimer of 543.33 PC(20:4) and 501.28 LPE (20:4)
TNFα TNFα Unknown
108
aHypoxia; bControl; cSequence represents amino acid composition of the peptide identified; d(Ac) represents a post-translation modification (PTM) by acetylation at b1; e(Ox) represents a PTM by oxidation; fHemogobin; gThe annotation (e.g. PC(16:1/2:0)) describes the number of carbon at-oms in the fatty acid chain with the number of double bonds. In this case, the fatty acid is com-posed of 16 carbons with one double bonds at the sn-1 position of the glycerol backbone, whereas 2:0 represents the 2 carbons with zero double bonds at sn-2 position; hLysophoshatidyl-choline; gLysophoshoethanolamine
997.61 (+1)
Dimer of 495.33 PC (16:0) and LPE (20:4)
TNFα TNFα
1504.85 Trimer of 501.28 LPE (20:4), 525.28 LPE (22:6/0:0) and 477.28 LPE (18:2/0:0)
TNFα TNFα
109
Among these related, responding compounds were cytoskeletal components including
fragments of vimentin and keratin-8 which were decreased in response to hypoxia. Several mole-
cules derived from thymosin B4 or B10 were also decreased with hypoxia and H/R. Wen et al.
found decreased circulating levels of thymosin B4 peptide in the serum of preeclamptic
women.48 This peptide’s downregulation may or may not be related to hypoxia, but the con-
sistent trend is suggestive.
Additionally, there was at least one inflammatory factor, 10,11-dihydro-12R-hydroxy-
leukotriene E4, was decreased in response to hypoxia and H/R. Interestingly, this same factor
was increased in response to TNFα as reported by others.37
In our studies, levels of 1,25 dihydroxy-vitamin D3 were significantly reduced in re-
sponse to hypoxia, consistent with trends seen in PE.39
Conditions applied to normal placenta changed features of fatty acid metabolism. In-
creased acyl-carnitines occurred with hypoxia and H/R. Similar changes have been reported in
hypoxic tissues and in PE.42
There were several changes in PCs and PEs. There were decreased PCs with hypoxia,
consistent with published studies.43, 44 In contrast there were increases in several lyso-PCs and
lyso-PEts with H/R or TNFα, also consistent with previous reports.45, 46 Three markers increased
by TNFα were dimers or a trimer of PCs and PEts. The PEt common to all three differentially
expressed molecules was lyso-PEt (20:4), also increased in TNFα-exposed placental explants.
Small amounts of lipid dimers may be formed in the ESI source49, 50 although other data suggests
that they are also present endogenously.51
110
Not all statistically significant markers were identified for several reasons: 1) The abun-
dance was too low for adequate fragment coverage. 2) High charge states of peptides make iden-
tification exceptionally difficult. This accompanied with their low initial abundance, post-transla-
tional modifications, peak overlap left fragment series incomplete or uninterpretable. 3) Having
little or no fragmentation precluded classification or component identification. Some fragments
were unrecognized in the current, limited lipid databases. For these, exact mass studies were per-
formed to predict elemental composition.
The second hypothesis in our studies was that there would be changes in LMW compo-
nents of placenta unique to each of the three applied PE conditions. Even among the most re-
sponsive molecules, there was some degree of overlap between control and stressed placenta and
between the three conditions also. The Venn diagram in Figure 3.3 represents the extent of over-
lap. To address the uniqueness to each PE condition, combinations of significantly different mol-
ecules were statistically modeled using LASSO. This resulted in several combinations able to
distinguish each abnormal condition imposed uniquely, a stress-specific molecular signature.
This satisfied our second hypothesis and should allow for evaluation of PE placenta to determine
their exposure to a particular stress.
In conclusion, these studies demonstrated the ability of a global ‘omics’ method, focused
on low molecular weight, low abundance species, to track the placenta’s response to abnormali-
ties considered participatory in PE and to develop signatures for each. These appear to provide
insights in pathology and may allow for identifying underlying pathology in PE placenta.
111
Figure 3.3. Venn diagram to represent the overlap in biomarkers between the three PE conditions applied on placental explants.
112
3.6 References
1. Ananth, C. V.; Keyes, K. M.; Wapner, R. J., Pre-eclampsia rates in the United States,
1980-2010: age-period-cohort analysis. Brit. Med. J. 2013, 347 (7937), 1-14.
2. Zeng, B.; Kwak-Kim, J.; Liu, Y.; Liao, A.-H., Treg Cells Are Negatively Correlated with
Increased Memory B Cells in Pre-eclampsia While Maintaining Suppressive Function on
Autologous B-Cell Proliferation. Am. J. Reprod. Immunol. 2013, 70 (6), 454-463.
3. Singh, H. J., Pre-Eclampsia: Is It All In The Placenta? Malays. J. Med. Sci. 2009, 16 (1), 7-
15.
4. Wang, Y.; Alexander, J. S., Placental pathophysiology in preeclampsia. Pathophysiology
2000, 6 (4), 261-270.
5. Soleymanlou, N.; Jurisica, I.; Nevo, O.; Ietta, F.; Zhang, X.; Zamudio, S.; Post, M.;
Caniggia, I., Molecular Evidence of Placental Hypoxia in Preeclampsia. J. Clin.
Endocrinol. Metab. 2005, 90 (7), 4299-4308.
6. Hung, T.-H.; Burton, G. J., Hypoxia and Reoxygenation: a Possible Mechanism for
Placental Oxidative Stress in Preeclampsia. Taiwan. J. Obstet. Gynecol. 2006, 45 (3), 189-
200.
7. Conrad, K. P.; Benyo, D. F., Placental cytokines and the pathogenesis of preeclampsia.
Am. J. Reprod. Immunol. 1997, 37 (3), 240-9.
8. Powe, C. E.; Levine, R. J.; Karumanchi, S. A., Preeclampsia, a disease of the maternal
endothelium: the role of anti-angiogenic factors and implications for later cardiovascular
disease. Circulation 2011, 123 (24), 2856-2869.
9. Robinson, J. M.; Vandre, D. D.; Ackerman, W. E. t., Placental proteomics: a shortcut to
biological insight. Placenta 2009, 30 (9), 13.
113
10. Centlow, M.; Hansson, S. R.; Welinder, C., Differential proteome analysis of the
preeclamptic placenta using optimized protein extraction. Biomed. Res. Int. 2009, 2010.
11. Gauster, M.; Siwetz, M.; Orendi, K.; Moser, G.; Desoye, G.; Huppertz, B., Caspases rather
than calpains mediate remodelling of the fodrin skeleton during human placental
trophoblast fusion. Cell Death Differ. 2009, 17 (2), 336-345.
12. Dunn, W. B.; Brown, M.; Worton, S. A.; Crocker, I. P.; Broadhurst, D.; Horgan, R.;
Kenny, L. C.; Baker, P. N.; Kell, D. B.; Heazell, A. E., Changes in the metabolic footprint
of placental explant-conditioned culture medium identifies metabolic disturbances related
to hypoxia and pre-eclampsia. Placenta 2009, 30 (11), 974-80.
13. Wagner, P. K.; Otomo, A.; Christians, J. K., Regulation of pregnancy-associated plasma
protein A2 (PAPPA2) in a human placental trophoblast cell line (BeWo). Reprod. Biol.
Endocrinol. 2011, 9, 48.
14. Ma, J.; Esplin, M. S.; Adair, C. D.; Mason, L. A.; Graves, S. W., Increasing Evidence for
and Regulation of a Human Placental Endogenous Digitalis-Like Factor. Reprod. Sci.
2012, 19 (4), 437-448.
15. Kedia, K.; Nichols, C. A.; Thulin, C. D.; Graves, S. W., Novel "omics" approach for study
of low-abundance, low-molecular-weight components of a complex biological tissue:
regional differences between chorionic and basal plates of the human placenta. Anal.
Bioanal. Chem. 2015, 407 (28), 8543-56.
16. Friedman, J.; Hastie, T.; Tibshirani, R., Regularization Paths for Generalized Linear
Models via Coordinate Descent. Journal of statistical software 2010, 33 (1), 1-22.
114
17. Harrell, F. E., Jr.; Lee, K. L.; Mark, D. B., Multivariable prognostic models: issues in
developing models, evaluating assumptions and adequacy, and measuring and reducing
errors. Stat. Med. 1996, 15 (4), 361-87.
18. Hand, D.; Till, R., A Simple Generalisation of the Area Under the ROC Curve for Multiple
Class Classification Problems. Machine Learning 2001, 45 (2), 171-186.
19. Brier, G., Verification of Forecasts expressed in terms of probability. Monthly Weather
Review 1950, 78 (1), 1-3.
20. Efron, B.; Tibshirani, R., Improvements on Cross-Validation: The .632+ Bootstrap
Method. J Am. Stat. Assoc. 1997, 92 (438), 548-560.
21. Slominski, A.; Semak, I.; Zjawiony, J.; Wortsman, J.; Li, W.; Szczesniewski, A.; Tuckey,
R. C., The cytochrome P450scc system opens an alternate pathway of vitamin D3
metabolism. The FEBS journal 2005, 272 (16), 4080-90.
22. Hevko, J. M.; Murphy, R. C., Electrospray ionization and tandem mass spectrometry of
cysteinyl eicosanoids: leukotriene C4 and FOG7. J. Am. Soc. Mass Spectrom. 2001, 12 (7),
763-71.
23. Orendi, K.; Kivity, V.; Sammar, M.; Grimpel, Y.; Gonen, R.; Meiri, H.; Lubzens, E.;
Huppertz, B., Placental and trophoblastic in vitro models to study preventive and
therapeutic agents for preeclampsia. Placenta 2011, 2011 (32), 023.
24. Gharesi-Fard, B.; Zolghadri, J.; Kamali-Sarvestani, E., Proteome differences of placenta
between pre-eclampsia and normal pregnancy. Placenta 2010, 31 (2), 121-5.
25. Ma, K.; Jin, H.; Hu, R.; Xiong, Y.; Zhou, S.; Ting, P.; Cheng, Y.; Yang, Y.; Yang, P.; Li,
X., A proteomic analysis of placental trophoblastic cells in preeclampsia-eclampsia. Cell
Biochem. Biophys. 2014, 69 (2), 247-58.
115
26. Leanos-Miranda, A.; Campos-Galicia, I.; Isordia-Salas, I.; Rivera-Leanos, R.; Romero-
Arauz, J. F.; Ayala-Mendez, J. A.; Ulloa-Aguirre, A., Changes in circulating
concentrations of soluble fms-like tyrosine kinase-1 and placental growth factor measured
by automated electrochemiluminescence immunoassays methods are predictors of
preeclampsia. J. Hypertens. 2012, 30 (11), 2173-81.
27. Heydarian, M.; McCaffrey, T.; Florea, L.; Yang, Z.; Ross, M. M.; Zhou, W.; Maynard, S.
E., Novel Splice Variants of sFlt1 are Upregulated in Preeclampsia. Placenta 2009, 30 (3),
250-255.
28. Gygi, S. P.; Rist, B.; Griffin, T. J.; Eng, J.; Aebersold, R., Proteome analysis of low-
abundance proteins using multidimensional chromatography and isotope-coded affinity
tags. J. Proteome Res. 2002, 1 (1), 47-54.
29. Jaquemar, D.; Kupriyanov, S.; Wankell, M.; Avis, J.; Benirschke, K.; Baribault, H.;
Oshima, R. G., Keratin 8 protection of placental barrier function. J. Cell Biol. 2003, 161
(4), 749-56.
30. Leloup, C.; Magnan, C.; Benani, A.; Bonnet, E.; Alquier, T.; Offer, G.; Carriere, A.;
Periquet, A.; Fernandez, Y.; Ktorza, A.; Casteilla, L.; Penicaud, L., Mitochondrial reactive
oxygen species are required for hypothalamic glucose sensing. Diabetes 2006, 55 (7),
2084-90.
31. Ahenkorah, J.; Hottor, B.; Byrne, S.; Bosio, P.; Ockleford, C. D., Immunofluorescence
confocal laser scanning microscopy and immuno-electron microscopic identification of
keratins in human materno-foetal interaction zone. J. Cell Mol. Med. 2009, 13 (4), 735-48.
116
32. Crockford, D.; Turjman, N.; Allan, C.; Angel, J., Thymosin beta4: structure, function, and
biological properties supporting current and future clinical applications. Ann. N. Y. Acad.
Sci. 2010, 1194, 179-89.
33. Moon, E. Y.; Im, Y. S.; Ryu, Y. K.; Kang, J. H., Actin-sequestering protein, thymosin
beta-4, is a novel hypoxia responsive regulator. Clin. Exp. Metastasis 2010, 27 (8), 601-9.
34. Young, J. D.; Lawrence, A. J.; MacLean, A. G.; Leung, B. P.; McInnes, I. B.; Canas, B.;
Pappin, D. J.; Stevenson, R. D., Thymosin beta 4 sulfoxide is an anti-inflammatory agent
generated by monocytes in the presence of glucocorticoids. Nat. Med. 1999, 5 (12), 1424-
7.
35. Ulrich, H. D., SUMO teams up with ubiquitin to manage hypoxia. Cell 2007, 131 (3), 446-
7.
36. Wennergren, G., Inflammatory mediators in blood and urine. Paediatr. Respir. Rev. 2000,
1 (3), 259-65.
37. Huber, M.; Beutler, B.; Keppler, D., Tumor necrosis factor alpha stimulates leukotriene
production in vivo. Eur. J. Immunol. 1988, 18 (12), 2085-8.
38. Zehnder, D.; Evans, K. N.; Kilby, M. D.; Bulmer, J. N.; Innes, B. A.; Stewart, P. M.;
Hewison, M., The ontogeny of 25-hydroxyvitamin D(3) 1alpha-hydroxylase expression in
human placenta and decidua. Am. J. Pathol. 2002, 161 (1), 105-14.
39. Shin, J. S.; Choi, M. Y.; Longtine, M. S.; Nelson, D. M., Vitamin D effects on pregnancy
and the placenta. Placenta 2010, 31 (12), 1027-34.
40. Wengrowski, A. M.; Kuzmiak-Glancy, S.; Jaimes, R., 3rd; Kay, M. W., NADH changes
during hypoxia, ischemia, and increased work differ between isolated heart preparations.
Am. J. Physiol. Heart Circ. Physiol. 2014, 306 (4), H529-37.
117
41. Meyburg, J.; Schulze, A.; Kohlmueller, D.; Linderkamp, O.; Mayatepek, E., Postnatal
Changes in Neonatal Acylcarnitine Profile. Pediatr. Res. 2001, 49 (1), 125-129.
42. Illsinger, S.; Janzen, N.; Sander, S.; Schmidt, K. H.; Bednarczyk, J.; Mallunat, L.; Bode, J.;
Hagebolling, F.; Hoy, L.; Lucke, T.; Hass, R.; Das, A. M., Preeclampsia and HELLP
syndrome: impaired mitochondrial function in umbilical endothelial cells. Reprod. Sci.
2010, 17 (3), 219-26.
43. Hatch, G. M.; Choy, P. C., Effect of hypoxia on phosphatidylcholine biosynthesis in the
isolated hamster heart. Biochem. J. 1990, 268 (1), 47-54.
44. Choy, P. C.; Chan, M.; Hatch, G.; Man, R. Y., Phosphatidylcholine metabolism in
ischemic and hypoxic hearts. Mol. Cell. Biochem. 1992, 116 (1-2), 53-8.
45. Ferrari, R. S.; Andrade, C., Oxidative Stress and Lung Ischemia-Reperfusion Injury. Oxid.
Med. Cell. Longev. 2015, 2015, 14.
46. Robinson, B. S.; Hii, C. S.; Poulos, A.; Ferrante, A., Effect of tumor necrosis factor-alpha
on the metabolism of arachidonic acid in human neutrophils. J. Lipid Res. 1996, 37 (6),
1234-45.
47. Jendryczko, A.; Drozdz, M., Increased placental phospholipase A2 activities in pre-
eclampsia. Zentralbl. Gynakol. 1990, 112 (14), 889-91.
48. Wen, Q.; Liu, L. Y.; Yang, T.; Alev, C.; Wu, S.; Stevenson, D. K.; Sheng, G.; Butte, A. J.;
Ling, X. B., Peptidomic Identification of Serum Peptides Diagnosing Preeclampsia. PLoS
One 2013, 8 (6), e65571.
49. Brady, J. J.; Judge, E. J.; Levis, R. J., Analysis of amphiphilic lipids and hydrophobic
proteins using nonresonant femtosecond laser vaporization with electrospray post-
ionization. J. Am. Soc. Mass Spectrom. 2011, 22 (4), 762-72.
118
50. Eibisch, M.; Zellmer, S.; Gebhardt, R.; Suss, R.; Fuchs, B.; Schiller, J.,
Phosphatidylcholine dimers can be easily misinterpreted as cardiolipins in complex lipid
mixtures: a matrix-assisted laser desorption/ionization time-of-flight mass spectrometric
study of lipids from hepatocytes. Rapid Commun. Mass Spectrom. 2011, 25 (18), 2619-26.
51. Anand, S.; Bench Alvarez, T. M.; Johnson, W. E.; Esplin, M. S.; Merrell, K.; Porter, T. F.;
Graves, S. W., Serum biomarkers predictive of pre-eclampsia. Biomark. Med. 2015, 9 (6),
563-75.
119
4 Targeted tissue ‘omics’ analysis comparing markers changed in culture in response to
proposed mediators of preeclampsia to those markers in actual preeclamptic placen-
tae
4.1 Abstract
Preeclampsia, a common disorder of pregnancy is one of the leading causes of maternal
and fetal morbidity and mortality. Globally, it effects 2-10% of pregnancies annually. The pla-
centa is believed to be the central feature in the pathogenesis of this disease. Several abnormali-
ties have been proposed to explain the cause of preeclampsia, the most common ones being pla-
cental hypoxia, oxidative stress to placenta and an increase in placental exposure to pro-inflam-
matory cytokines. As described in Chapter 3, these abnormalities were imposed on healthy pla-
cental explants which resulted in significant differential expression of 146 markers in total. Of
these, 75 changed with introduction of hypoxia, whereas 23 and 48 were differentially expressed
in response to oxidative stress and tumor necrosis factor-alpha. We sought to identify which, if
any of these changes were also expressed in actual PE pregnancies. By using ANCOVA and
linear regression models as joint criteria for significance, 4 out of 146 biomolecules demon-
strated significant differences between preeclamptic placenta and placentae collected from
healthy controls. In the initial study (Chapter 3), 2 of these changes were brought about by tumor
necrosis factor-alpha (m/z 461.06 (z=+1) and m/z 476.24 (z=+1)), 1 due to hypoxia-reoxygena-
tion (m/z 426.35 (z=+1)) and one with hypoxia (m/z 649.49 (z=+1)). Biomarkers with m/z’s
476.24 and 426.35 were identified as a phosphatidylcholine and an acylcarnitine respectively,
while the identities of 2 were unknown. Previous research has demonstrated an increase in the
abundances of phospholipids as well as acylcarnitines in preeclamptic placentae.
120
4.2 Introduction
Preeclampsia (PE) is a complex disorder of pregnancy characterized by hypertension
(>140/90 mmHg) and proteinuria (>300 mg/24 hr).1 It is the second leading cause of maternal
and fetal morbidity following hemorrhage.2 In severe cases it leads to eclampsia: a condition
characterized by grand mal seizures not associated with any previously known abnormal neuro-
logic activity.3 PE is believed to affect 2-10% of pregnancies worldwide.3 Each year around ten
million women develop PE globally and over 75,000 of these succumb to it.4 This syndrome is
also responsible for the deaths of over 500,000 infants annually.5 Due to the absence of any diag-
nostic tests, PE is most commonly diagnosed by observing and confirming symptoms such as
high blood pressure and proteinuria. There are no therapeutic remedies to prevent or treat PE.
PE is a condition of abnormal placental function as highlighted by the fact that the symp-
toms of PE are reversed once the placenta is removed from the body by either vaginal delivery or
surgery.6 In addition PE has been shown to develop in molar pregnancies as well, thus confirm-
ing its involvement in the development of this disease.6
A number of factors are believed to contribute to the pathogenesis of PE. The most widely
accepted ones include placental hypoxia,7 oxidative stress of the placenta,8 and/or exposure to
increased pro-inflammatory mediators.9 The aim of our research was to identify which, if any of
these hypothesized causative pathways play an important role in the etiology of this unexplained
syndrome as identified by unique expression patterns of less abundant peptides, low molecular
weight proteins, and potentially other biomolecules. To accomplish this goal two studies were
performed. In the first study, healthy human placental explants were cultured under the proposed
mediators of PE namely, hypoxia, oxidative stress and exposure to inflammatory cytokines, ac-
companied by a control group where no provocative agent was introduced and the explants were
121
cultured in normal conditions of oxygen. These explants were incubated for a total of 48 hrs in
their respective environment. They were then processed separately and analyzed by MS. The MS
data was analyzed to study differentially expressed molecules between the comparison groups and
therefore develop a signature of molecules specific to each abnormality. The second study, re-
ported here, involves peptidomic analysis of those biomolecules demonstrating statistically signif-
icant differences between ‘stressed’ and ‘unstressed’ placenta in placental tissue collected from
established preeclamptic pregnancies and comparing them with healthy placentas collected and
processed identically.
4.3 Methods
4.3.1 Tissue collection
The Internal Review Boards of Erlanger Hospital, University of Tennessee, Chattanooga
and Brigham Young University approved of these studies. All the specimens were obtained from
Erlanger Hospital, University of Tennessee, Chattanooga. Placentas from C-sections and vaginal
deliveries were collected, dissected at the intervillous region and flash frozen using liquid nitro-
gen. In total 26 placental tissues were used in this study, 13 from preeclamptic pregnancies, and
13 from uncomplicated pregnancies. Information regarding gestational age (GA) is provided (Ta-
ble 4.1) and placentae from pregnancies with pre-existing conditions such as hypertension and
diabetes were excluded from the study.
122
Table 4.1. Mean and standard deviation of GA for preeclamptic and healthy pregnan-cies
PE (n=13) Controls (n=13) 32.7±5 40.0±1
123
4.3.2 Specimen processing
A block of placental tissue obtained approximately halfway between the cord and the rim.
The chorionic plate was dissected away and a ~ 1cm region of the remaining tissue halfway be-
tween chorion interface and the basal plate was obtained. This block representing intervillous tis-
sue was shaved into thin slices using a stainless steel blade and regions with high concentrations
of blood vessels and fat deposits were avoided. The tissue slices were further minced and trans-
ferred to an autoclaved, sterile beaker. Multiple washes were performed using 1X Phosphate
buffered saline (PBS) to remove most of the external blood from the tissue. Washed tissue was
then transferred to a stack of autoclaved tissue paper to pat dry them. Following this, 300 mg of
dried tissue was then processed, which included tissue homogenization and acetonitrile precipita-
tion. Details of this procedure are described in Chapter 3, but briefly the weighed tissue was ho-
mogenized in the presence of 20 µL of protease inhibitor cocktail (P9599, Sigma -Aldrich) and
20 µL of 8.87 mM 1,10-phenanthroline (Sigma -Aldrich). Homogenate was resuspended in 3 mL
of 1X PBS, vortexed and centrifuged to remove cell debris. An aliquot of the supernatant was
then treated with two volumes of acetonitrile (2:1 v/v). This step is used to deplete high molecu-
lar weight (HMW), abundant proteins in order to prevent ion suppression during the MS analy-
sis, thus allowing the interrogation of a much larger set of less abundant species. A known, fixed
amount of the protein-depleted sample was then injected into cLC-ESI-MS.
4.3.3 Chromatographic separation
An Agilent 1260 Infinity Series HPLC pump system (Agilent Technologies, Karlsruhe,
Germany) was part of the LC-MS setup. A volume of 8 µL, containing an apparent protein con-
centration of 1 µg/µL were injected onto a 1 mm (16.2 µL) dry-packed MicroBore guard column
124
(IDEX Health and Science, Oak Harbor, WA), coupled to a 15 cm x 250 µm i.d. capillary col-
umn, slurry-packed in-house with POROS R1 reverse-phase medium (Applied Biosystems, Fos-
ter City, CA) with a mobile phase flow rate of 5 µL/min. Mobile phase A was an aqueous solu-
tion water/acetonitrile/formic acid (98:2:0.1, v/v/v) and mobile phase B was primarily organic
solution containing acetonitrile/water/formic acid (98:2:0.1, v/v/v). Information on the gradient
elution has been provided in Chapter 2.
4.3.4 Mass spectrometric analysis
The eluate from the column was introduced into the quadrupole (Q)-time-of-flight (TOF)
tandem mass spectrometer (Q-TOFMS) through an electrospray needle to allow in-line analysis
(Agilent Technologies 6530 Accurate-Mass Q-TOF/LC/MS). The electrospray needle was set at
3800 V and all the samples were run in the positive ion mode. Mass spectra were obtained over
the mass to charge ratio (m/z) range of 400–3000, collected in profile mode for each mass spec-
trum with an acquisition rate of 8 spectra/sec. Data was acquired using MassHunter Data Acqui-
sition B.06.00 and was analyzed using the instrument’s Qualitative MassHunter B.06.00 soft-
ware (Agilent Technologies). Other MS parameters were as follows: gas temperature: 300°C,
drying gas flow rate: 5 L/min, nebulizer pressure: 15 psi, fragmentor voltage: 175 V, skimmer
voltage: 65 V.
4.3.5 Data analysis
Time normalization, as described previously in Chapter 3, was achieved by defining 2 min
windows along the entire LC chromatogram. These reference time markers were used to align
elution windows across different runs. The 11 time markers identified for use in these studies of
125
human placenta are summarized in Results section. In total 146 markers were found to be differ-
entially expressed due to the imposed preeclamptic abnormalities to healthy human placental ex-
plants. In order to determine the continued statistical significance of these markers in actual
preeclamptic pregnancies, MS data from each of the 11 time windows was exported and the in-
tensities of the 146 targeted individual markers of interest were computed. This analysis was per-
formed by using Agilent’s MassHunter Qualitative Analysis B.06.00 software. Following this,
statistical tests were performed to evaluate significance of individual markers in this study.
4.3.6 Statistical analysis
Preliminary statistical analysis revealed 8 biomarkers out of 146, continuing to express sta-
tistically significant differences between the preeclamptic and control placenta. Most of the pla-
centa from preeclamptic patients were delivered earlier than those from women with normal
pregnancies. And because the intensity of the markers may be influenced by GA of the placenta,
approaches were taken to account for the effect of GA in searching for statistically significant
markers. Thus two more stringent statistical procedures were implemented to negate for the ef-
fect of different GA. The first approach uses an analysis of covariance (ANCOVA) model, which
compares the average of the measured variables in two different classes (i.e. preeclamptic or con-
trol). The covariate in this model is the GA of the each of patients at delivery which is adjusted
out of the comparison of the average difference between control and preeclamptic deliveries.
The second approach used simple linear regression to regress the GA on the peak intensi-
ties for just the control samples and then to estimate a prediction interval for GAs in the range of
the preeclamptic samples. These predicted intervals are then compared with the observed peak
126
intensities from the preeclamptic samples. If there is no difference in the intensities of these vari-
ables between the controls and the preeclamptic deliveries, most of the observed variables should
fall within the prediction intervals for deliveries at that GA.
One observation was missing a GA value, so an estimate was obtained using multiple im-
putation by chained equations based on predictive mean matching.10
The two procedures were used because both have particular strengths and weaknesses as
analytical tools. The ANCOVA model is good because it directly compares the mean intensities
of the peaks between the controls and the preeclamptic placentae while also accounting for the
effect of GA. However, there is not much overlap between the GA of the two classes. Conse-
quently, the adjustment for GA relies heavily on those few pregnancies of control and
preeclamptic women that have similar GA and the assumption that the pattern seen for deliveries
at older GAs can be linearly adjusted to the younger GA. Furthermore, the 2 independent varia-
bles, class and GA, are correlated, violating a principal assumption of the model.
The regression model is good in that it directly models the effect of GA, which can be used
in detecting effects beyond the GA. These effects are attributed to the class. However, it was
necessary to extrapolate to lower GA, requiring that the assumption of a linear relationship be-
tween GA and peak intensity be valid. Despite the inherent weaknesses, if the peak intensities for
the controls and preeclamptic samples are sufficiently different to be deemed significant by both
procedures, we determine that such a marker has a significant relationship with preeclampsia. It
should be noted that, even with the precautions we take in accounting for GA, the observed sta-
tistical significance may be confounded by GA.
The ANCOVA model used for a specific peak is as follows: Yij=μ+γXi+τij+ϵij, where Yij is
the peak intensity for the jth placenta in the ith class, μ is the overall mean, γ is the effect of GA,
127
Xij is the GA for the jth placenta in the ith class, τi is the effect of GA for the ith class, and ϵij is the
residual of the jth placenta in the ith class. Due to the variability of the peak intensities and the
likely presence of outliers, the two largest and smallest peak intensities for both classes for each
peak were excluded from this analysis, leaving 9 samples per class. The class effect was calcu-
lated based on Type III sums of squares, and any marker with a class effect p-value less than 0.1
was counted as significant.
The regression model used is as follows: Yi1=β0+β1Xi1+ϵi1, where Yi1 is the peak intensity
for the ith individual in the control class, β0 is the intercept, β1 is the estimated coefficient for GA,
Xi1 is the GA for the ith individual in the control class, and ϵi1 is the residual for the ith individual
in the control class. If the p-value for β1 was non-significant (p>0.1), a model without the GA ef-
fect (i.e. a model including just the intercept) was built, and this reduced model was compared
against the model that included the GA effect using an F test; all tests resulted in non-significant
p-values, indicating that there was not enough evidence to suggest that, for the given peak, the
model including GA as an independent variable was better than the model without any adjust-
ment of GA. Despite the apparent non-significance of GA in many of the models, it was included
in some as a conservative measure for adjusting small effects on the measured intensities. Using
the selected model, prediction intervals were then estimated spanning the range of all GA. To see
if the peak values of the preeclamptic samples followed the same trend of GA as the control sam-
ples, the (GA, peak intensity) points were plotted for all preeclamptic samples. Assuming that
the preeclamptic samples follow the same pattern as the control samples, it is expected, based on
the 95% prediction interval, that 95% of the observed preeclamptic points should be contained
within the interval; if significantly fewer than 95% are inside the interval, there’s evidence that
there is an effect of preeclampsia beyond the GA effect. Probabilities for the number of points
128
outside the interval based on the binomial distribution were calculated, and markers where the
probability was less than 0.1 were counted as significant.
4.4 Results
4.4.1 Chromatographic time alignment
The 11 time normalization peaks are summarized by their m/z, charge state and approxi-
mate elution time as follows: 695.09 (z=+4) 17 min, 827.28 (z=+6) 19 min, 474.2 (z=+1) 21 min,
672.36 (z=+3) 23 min, 686.46 (z=+1) 25 min, 1009.05 (z=+1) 27 min, 616.16 (z=+1) 32 min,
526.28 (z=+1) 34 min, 524.36 (z=+1) 37 min, 650.43 (z=+1) 39 min, 675.53 (z=+1) 41 min.
4.4.2 Evaluation of 146 biomarkers that demonstrated differences under PE conditions in PE
placenta
Previously, a total of 146 biomarkers were found to be differentially expressed in healthy
placental explants maintained under PE conditions (Chapter 3). Of these 75 changed in response
to hypoxia; 23 to hypoxia-reoxygenation (an oxidative stress) and 48 in response to TNFα expo-
sure.
On average, placenta collected from actual preeclamptic pregnancies had earlier GAs than
those collected from healthy controls. Since there was a correlation between earlier GA and
preeclamptic placenta, we undertook measures to account for the potentially confounding effect
using 2 statistical models. The ANCOVA model identified 9 significant markers, and the linear
regression model identified 32 significant markers. However, only 4 of these markers exhibited
significant differences when analyzed by both tests, providing the strongest evidence to suggest
129
that these markers have significantly different intensities between normal and preeclamptic pla-
centa.
These models found 4 biomarkers with significant differences in abundance between PE
placenta and controls. These markers are described in Table 4.2.
130
Table 4.2. Biomarker with statistical significant differences between PE placenta and healthy controls.
m/z z Higher in p-value (ANCOVA) 461.06 +1 TNFα 0.0487 476.24 +1 TNFα 0.0368 426.35 +1 Hypoxia-reoxygenation 0.0471 649.49 +1 Hypoxia 0.0735
131
All these 4 biomarkers showed an increased expression in preeclamptic placentae. As can
be compared to Chapter 3, 2 of these markers were more abundant in placental explants cultured
with TNFα, 1 more abundant in explants cultured under hypoxia and the remaining 1 in explants
cultured under conditions producing oxidative stress (hypoxia-reoxygenation). Details of these
markers and related information as gathered from the previous study is given in Table 4.3.
132
Table 4.3. Statistically significant markers with their identities and average elution time in the total ion chromatogram
m/z Z ID Elemental composition Average elution time (min)
461.06 +1 Unknown Unknown 18.22±0.3a
476.24 +1 Phosphatidylcholine C20H40NO8NaP 26.47±0.7 426.35 +1 Fatty acyl carnitine
O-oleoylcarnitine C25H48NO4 36.87±0.7
649.49 +1 Unknown Unknown 40.14±0.1 aStandard deviation of elution time.
133
As can be seen in the table, 2 of the 4 markers have been identified. The biomarker with
m/z 426.35 was identified as an acylcarnintine with a fatty acid chain containing 18 carbons and
one double bond (C18:1, oleoylcarnitine). Identification of acylcarnitine was based on the pro-
duction of signature product ion peaks in the MS/MS spectrum of the fragmented precursor. The
most common ones are at m/z 60, 85 and 144 (Figure 4.1). Elemental composition was con-
firmed by performing exact mass studies as well by carrying out a database search based on the
MS1 precursor mass (http://www.lipidmaps.org/). For the second biomarker (m/z 476.24), a peak
at m/z 184.07 was seen which indicated the presence of a phosphocholine headgroup. Exact mass
studies accompanied by use of an elemental composition calculator revealed the molecular for-
mula as C20H40NO8NaP, which represents a sodiated PC (M+Na of m/z 454.24). For one of the
unknown markers (m/z 461.01), we observed 3 very abundant product ion peaks at [M+H-180],
[M+H-44] and [M+H-180-44]. Though we saw a pattern of neutral losses as well as very abun-
dant product ions, we were still unable to determine the identity of this markers due to absence of
any related information in the database or in the literature.
134
Figure 4.0.1. MS/MS spectra of identified acylcarnitine (m/z 426.35; z=+1))
135
4.5 Discussion
PE still remains a disease with unknown cause. Several mediators have been proposed to
be involved in the pathogenesis of this disease.7-9 Several research groups have performed tar-
geted or untargeted studies to identify features that differ in diseased and healthy placentae. Tar-
geted approaches were performed to monitor a single or a few biomolecules which may play a
role in the pathogenesis of PE.11-13 Untargeted approaches have been most commonly used in
terms of performing a more extensive analysis on the placenta from diseased and healthy tis-
sue.14, 15 Most of the proteomics techniques that have been performed on placental tissue to dis-
sect features of PE have used gel electrophoretic approaches to separate proteins based on their
charge and mass, which is followed by mass spectrometry.16 These approaches, including 2-
DGE, are most appropriate for the analysis of abundant HMW proteins.17 The importance of less
abundant, LMW biomolecules has been well described in literature.18 This fraction of the protein
repertoire as well as other LMW metabolites plays important part in the pathophysiology of dis-
eases.
Tissue-proteomics approach developed in our lab allowed us to explore 7000-8000 less
abundant LMW species in a single MS run. Of these many species, a previous study conducted
by our group observed changes in 146 biomolecules in human placental explants under
preeclamptic condition. As the most commonly hypothesized abnormalities in PE are placenta
hypoxia, oxidative stress and accumulation of pro-inflammatory cytokines, we investigated
which biomolecules showed differential abundances as a consequence of this exposure in healthy
placental explants. Details of this work can be found in Chapter 3. Out of the 146 biomarkers,
136
hypoxia bought changes in 75 of these biomarkers, while 23 were due to effects produced by hy-
poxia-reoxygentaion injury and 48 due to TNFα. The aim of this study was to seek which, if any
of these differences are also revealed in actual preeclamptic placentae.
By submitting all these markers to the two statistical models described in methods section
we were able to find 4 biomarkers with differential expression in PE placenta when compared to
healthy controls common to both analyses. We note that because of the discrepancy in GA be-
tween PE and control women, it makes it more difficult to determine whether any difference we
found in any marker was due to the pathology of the disease or a gestational age related change.
When we looked at the controls, there was no association between gestational age and biomarker
abundance for any of these markers, suggesting perhaps that the changes observed were due to
disease rather than GA. However, the GA range was limited in the control group. The two indi-
vidual analyses that corrected for GA found several (9 and 32 respectively) markers that ap-
peared different even after considering the contribution of GA. Nonetheless, those markers that
were found statistically different in all analyses, suggested most strongly that they represent real
disease related changes.
Two of these markers with m/z 461.06 (z=+1, p=0.04), 476.24 (z=+1, p=0.03), had been
elevated in placental explants treated with TNFα, one of the markers with m/z 426.35 (z=+1,
p=0.04) was changed with hypoxia/reoxygenation and the marker m/z 649.49 (z=+1, p=0.07)
was changed with hypoxia. The biomarkers with m/z values of 426.35 and 476.24 were identi-
fied as an acycarnitine and phosphatidylcholine respectively, while the identities of the remain-
ing 2 were unable to be identified.
Previous research has shown that the composition of PC and PE in human placenta is 45%
and 29%-30%, respectively.19 Several research groups have indicated a higher expression of
137
phospholipids in PE. It has been believed that endothelial dysfunction, a characteristic of PE, in
part is due to an elevated level of lipids in the placenta. This included increases in the decidual
basalis of placenta as well as in the placenta proper.19, 20
In this study, we observed a higher abundance of PC in particular, for both placental tissue
exposed to PE conditions as well as in actual PE placenta. The exact mechanism and the conse-
quences for this change remain unknown, but several theories are proposed to provide an expla-
nation to this increased abundance. One study suggested an activation of the coagulation cascade
caused by an increased concentration of phospholipids, as it has been hypothesized that PE like
symptoms are developed in the presence of hypercoagulation in the placental circulation.19 In-
creased disseminated small clot formation and depleted platelet numbers are common features of
PE. In another study by Omatsu et al., a PE like state with hypercoagulation was developed by
injecting phosphocholine/phosphoserine macrovesicles into pregnant mice.21 Another theory
suggests a decreased protein expression of ATP-binding cassette (ABC) transporters, in particu-
lar ABCA1 in the STB of PE placental tissue. ABCA1 is expressed in placenta and is responsible
for phospholipid efflux. A decreased expression of these transporters might be responsible for an
accumulation of phospholipids in PE placenta.22
Fatty acids are considered to be the major metabolic fuel for placenta and are made availa-
ble to the mitochondria through a series of steps that begins with the production of acyl-CoA es-
ters.23 It is in the mitochondria that fatty acid oxidation with production of high energy molecules
takes place but the mitochondrial membrane is impermeable to them.24 Carnitine acts as a shuttle
for the fatty acids with transfer from their CoA-esters to form acylcarnitines which enter the in-
ner mitochondria where acyl-CoA esters are formed again and undergo β-oxidation to form ace-
tyl-CoA, NADH and FADH2.24 Acetyl-CoA enters the citric acid cycle to produce additional
138
NADH and FADH2 used in the electron transport chain (ETC) to generate ATP. Hypoxia-reoxy-
genation impairs mitochondrial ETC function,25 and as a consequence there is an increased accu-
mulation of acylcarnitines as they remain unused.26 Studies have reported an increase in acyl-
carnitines in PE.27, 28 Additionally, Thiele et al found a 50% increase in carnitine concentration in
maternal blood in preeclamptic women. These excess carnitines may also cause tissue damage
contributing to PE.29
Insufficient spiral artery remodeling has been proposed to be one of the initial events lead-
ing to PE, causing a reduced blood flow to the placenta. It has been proposed that this defect in
remodeling can cause placental hypoxia30 or hypoxia accompanied with an accumulation of
TNFα,31 or an accumulation of TNFα due to placental hypoxia9 as well as oxidative stress in pla-
centa.8 This study did not necessarily answer which abnormality, in particular, might be the rea-
son for the physiological changes seen in PE, but it did provide information in terms of which
biomolecules demonstrate differential abundance in placental tissue under these conditions as
well in actual preeclamptic placentae.
One of the reasons for seeing significant differences only in 4 biomolecules and not more,
might be a feature of the timing of pathophysiological stages in PE. The proposed abnormalities
initiating PE may well occur early in pregnancy, mostly during the time of placentation, but may
not persist through the next several weeks of pregnancy. In the explant culture study that has
been described in Chapter 3, PE abnormalities were introduced in full term healthy placental ex-
plants, and only for a period of 48 hrs. It is very likely that in fully established PE, markers tied
to early changes have already occurred and may then be followed by very different changes. It is
also possible that the 48-hr exposure window does not capture the long term sequelae of a much
139
longer exposure. While it seems less likely, given that these are not full length proteins but actu-
ally relatively stable lipids studied here, there may be differences due to the PE and control pla-
centa studied here having been frozen. Also, the mix of both Caesarian and vaginal deliveries in
this study may well have increased variability in the data. Finally, it is possible that the actual
causes or mediators of active PE are not those proposed by other investigators and are more in-
ferred than real. The 4 markers we did see were distributed across all 3 possible pathologic in-
sults considered here rather than a single pathology.
4.6 Conclusion
To conclude, we have investigated which less abundant, LMW biomolecules are differen-
tially expressed in healthy human placental tissue when exposed to proposed abnormalities of
PE, and which of these differentially expressed biomarkers are also different in placentae col-
lected from actual PE pregnancies when compared to heathy controls processed and analyzed
identically.
140
4.7 References
1. Brown, M. A.; Lindheimer, M. D.; de Swiet, M.; Assche, A. V.; Moutquin, J.-M., The
Classification and Diagnosis of the Hypertensive Disorders of Pregnancy: Statement from
the International Society for the Study of Hypertension in Pregnancy (ISSHP). Hypertens.
Pregnancy 2001, 20 (1), ix-xiv.
2. Walsh, C. A.; Baxi, L. V., Mean arterial pressure and prediction of pre-eclampsia. BMJ :
Brit. Med. J. 2008, 336 (7653), 1079-1080.
3. Osungbade, K. O.; Ige, O. K., Public health perspectives of preeclampsia in developing
countries: implication for health system strengthening. J. Pregnancy 2011, 2011, 481095.
4. Zhou, C. C.; Irani, R. A.; Dai, Y.; Blackwell, S. C.; Hicks, M. J.; Ramin, S. M.; Kellems,
R. E.; Xia, Y., Autoantibody-mediated IL-6-dependent endothelin-1 elevation underlies
pathogenesis in a mouse model of preeclampsia. J. Immunol. 2011, 186 (10), 6024-34.
5. English, F. A.; Kenny, L. C.; McCarthy, F. P., Risk factors and effective management of
preeclampsia. Integr. Blood Press. Control 2015, 8, 7-12.
6. Roberts, J. M.; Escudero, C., The placenta in preeclampsia. Pregnancy Hypertens. 2012, 2
(2), 72-83.
7. Soleymanlou, N.; Jurisica, I.; Nevo, O.; Ietta, F.; Zhang, X.; Zamudio, S.; Post, M.;
Caniggia, I., Molecular Evidence of Placental Hypoxia in Preeclampsia. J. Clin.
Endocrinol. Metab. 2005, 90 (7), 4299-4308.
8. Hung, T.-H.; Burton, G. J., Hypoxia and Reoxygenation: a Possible Mechanism for
Placental Oxidative Stress in Preeclampsia. Taiwan. J. Obstet. Gynecol. 2006, 45 (3), 189-
200.
141
9. Conrad, K. P.; Benyo, D. F., Placental cytokines and the pathogenesis of preeclampsia.
Am. J. Reprod. Immunol. 1997, 37 (3), 240-9.
10. Buuren, S. v.; Groothuis-Oudshoorn, K., mice: Multivariate Imputation by Chained
Equations in R. Journal of statistical software 2011, 45 (3), 67.
11. Gu, Y.; Lewis, D. F.; Wang, Y., Placental Productions and Expressions of Soluble
Endoglin, Soluble fms-Like Tyrosine Kinase Receptor-1, and Placental Growth Factor in
Normal and Preeclamptic Pregnancies. J. Clin. Endocrinol. Metab. 2008, 93 (1), 260-266.
12. Heydarian, M.; McCaffrey, T.; Florea, L.; Yang, Z.; Ross, M. M.; Zhou, W.; Maynard, S.
E., Novel Splice Variants of sFlt1 are Upregulated in Preeclampsia. Placenta 2009, 30 (3),
250-255.
13. Sano, M.; Matsumoto, M.; Terada, H.; Wang, H.; Kurihara, Y.; Wada, N.; Yamamoto, H.;
Kira, Y.; Tachibana, D.; Koyama, M., Increased annexin A2 expression in the placenta of
women with acute worsening of preeclampsia. Osaka City Med. J. 2014, 60 (2), 87-93.
14. Yang, J. I.; Kong, T. W.; Kim, H. S.; Kim, H. Y., The Proteomic Analysis of Human
Placenta with Pre-eclampsia and Normal Pregnancy. J. Korean Med. Sci. 2015, 30 (6),
770-778.
15. Baig, S.; Kothandaraman, N.; Manikandan, J.; Rong, L.; Ee, K. H.; Hill, J.; Lai, C. W.;
Tan, W. Y.; Yeoh, F.; Kale, A.; Su, L. L.; Biswas, A.; Vasoo, S.; Choolani, M., Proteomic
analysis of human placental syncytiotrophoblast microvesicles in preeclampsia. Clin.
Proteomics 2014, 11 (1), 40.
16. Wang, F.; Wang, L.; Xu, Z.; Liang, G., Identification and Analysis of Multi-Protein
Complexes in Placenta. PLoS One 2013, 8 (4), e62988.
142
17. Ma, K.; Jin, H.; Hu, R.; Xiong, Y.; Zhou, S.; Ting, P.; Cheng, Y.; Yang, Y.; Yang, P.; Li,
X., A proteomic analysis of placental trophoblastic cells in preeclampsia-eclampsia. Cell
Biochem. Biophys. 2014, 69 (2), 247-58.
18. Merrell, K.; Southwick, K.; Graves, S. W.; Esplin, M. S.; Lewis, N. E.; Thulin, C. D.,
Analysis of low-abundance, low-molecular-weight serum proteins using mass
spectrometry. J. Biomol. Tech. 2004, 15 (4), 238-48.
19. Huang, X.; Jain, A.; Baumann, M.; Körner, M.; Surbek, D.; Bütikofer, P.; Albrecht, C.,
Increased Placental Phospholipid Levels in Pre-Eclamptic Pregnancies. Int. J. Mol. Sci.
2013, 14 (2), 3487-3499.
20. Staff, A. C.; Ranheim, T.; Khoury, J.; Henriksen, T., Increased contents of phospholipids,
cholesterol, and lipid peroxides in decidua basalis in women with preeclampsia. Am. J.
Obstet. Gynecol. 1999, 180 (3 Pt 1), 587-92.
21. Omatsu, K.; Kobayashi, T.; Murakami, Y.; Suzuki, M.; Ohashi, R.; Sugimura, M.;
Kanayama, N., Phosphatidylserine/phosphatidylcholine microvesicles can induce
preeclampsia-like changes in pregnant mice. Semin. Thromb. Hemost. 2005, 31 (3), 314-
20.
22. Korner, M.; Wenger, F.; Nikitina, L.; Baumann, M.; Surbek, D.; Albrecht, C., PP141. The
lipid transporters ABCA1 and ABCG1 are differentially expressed in preeclamptic and
IUGR placentas. Pregnancy Hypertens. 2012, 2 (3), 315-6.
23. Shekhawat, P.; Bennett, M. J.; Sadovsky, Y.; Nelson, D. M.; Rakheja, D.; Strauss, A. W.,
Human placenta metabolizes fatty acids: implications for fetal fatty acid oxidation
disorders and maternal liver diseases. Am. J. Physiol. Endocrinol. Metab. 2003, 284 (6),
E1098-105.
143
24. Sharma, S.; Black, S. M., Carnitine homeostasis, mitochondrial function, and
cardiovascular disease. Drug discovery today. Disease mechanisms 2009, 6 (1-4), e31-e39.
25. Thu, V. T.; Kim, H.-K.; Long, L. T.; Lee, S.-R.; Hanh, T. M.; Ko, T. H.; Heo, H.-J.; Kim,
N.; Kim, S. H.; Ko, K. S.; Rhee, B. D.; Han, J., NecroX-5 prevents hypoxia/reoxygenation
injury by inhibiting the mitochondrial calcium uniporter. Cardiovasc. Res. 2012, 94 (2),
342-350.
26. Meyburg, J.; Schulze, A.; Kohlmueller, D.; Linderkamp, O.; Mayatepek, E., Postnatal
Changes in Neonatal Acylcarnitine Profile. Pediatr. Res. 2001, 49 (1), 125-129.
27. Illsinger, S.; Janzen, N.; Sander, S.; Schmidt, K. H.; Bednarczyk, J.; Mallunat, L.; Bode, J.;
Hagebolling, F.; Hoy, L.; Lucke, T.; Hass, R.; Das, A. M., Preeclampsia and HELLP
syndrome: impaired mitochondrial function in umbilical endothelial cells. Reprod. Sci.
2010, 17 (3), 219-26.
28. Oey, N. A.; den Boer, M. E.; Ruiter, J. P.; Wanders, R. J.; Duran, M.; Waterham, H. R.;
Boer, K.; van der Post, J. A.; Wijburg, F. A., High activity of fatty acid oxidation enzymes
in human placenta: implications for fetal-maternal disease. J. Inherit. Metab. Dis. 2003, 26
(4), 385-92.
29. Thiele, I. G. I.; Niezen-Koning, K. E.; van Gennip, A. H.; Aarnoudse, J. G., Increased
Plasma Carnitine Concentrations in Preeclampsia. Obstet. Gynecol. 2004, 103 (5, Part 1),
876-880.
30. Tal, R., The Role of Hypoxia and Hypoxia-Inducible Factor-1Alpha in Preeclampsia
Pathogenesis. Biol. Reprod. 2012, 87 (6), 134, 1-8.
144
31. Benyo, D. F.; Smarason, A.; Redman, C. W.; Sims, C.; Conrad, K. P., Expression of
inflammatory cytokines in placentas from women with preeclampsia. J. Clin. Endocrinol.
Metab. 2001, 86 (6), 2505-12.
145
5 Concluding remarks
The overall objective of this work was to investigate which of the several hypothesized
causal conditions or factors play a role in the etiology of PE as identified by the expression signa-
ture of the less abundant, low molecular weight proteins and other biomolecules. Placenta’s in-
volvement in the pathogenesis of PE remains unequivocal, which makes it an organ of great inter-
est in the study of PE.
5.1 Summary of Current research
5.1.1 Summary: Chapter 2
An approach to study lower molecular weight, lower abundance biomolecules constituents
in a tissue was developed in our lab that involved homogenization of tissue followed by acetonitrile
precipitation to remove abundant, high molecular weight proteins.1 Considering the cellular and
anatomic complexity of placenta, it was important to test the applicability of this method to study
placental features. In an attempt to do so, we investigated regional differences in human placenta
as a proof of concept for our approach. Due to the somewhat different cellular composition of the
chorionic plate and basal plate of placenta, these two regions have subtle differences in their ex-
pressed biomolecules. Such differences should ideally be detectable using our method if we are to
apply it to investigating biomolecules in preeclamptic placenta. In our study we found region spe-
cific, statistically significant differences in 16 LMW markers between these two sides. Of these,
12 species were identified and/or chemically characterized. Thus, we conclude that our method
146
will be capable of studying placental features and identifying even subtle differences existing in
diseased conditions.
5.1.2 Summary: Chapter 3
Once the placental tissue-homogenization approach was developed and optimized to be ap-
plied to human placenta, we moved forward to our second study which involved introducing pro-
posed mediators of PE e.g. hypoxia, oxidative stress and an increased concentration of TNFα, a
cytokine implicated in PE, to healthy human placental explants collected from elective C-section
from uncomplicated pregnancies. This study tested two hypothesis. The first was that there will be
significant changes in abundances of some biomolecules as a result of preeclamptic conditions
applied to placenta and second that each condition or abnormality will produce a unique set of
changes, i.e. a molecular signature. In total, 146 biomolecules had statistically significant changes
when combining all three comparison groups, the breakdown being: 75 changes due to hypoxia,
23 due to hypoxia-reoxygenation and 48 due to TNFα. Using tandem MS, 45% of 146 molecules
were identified and/or chemically characterized. To develop signatures specific to each stressor,
the Lasso statistical model identified 16 markers which were able to well segregate all 4 treatment
groups.
5.1.3 Summary: Chapter 4
Placentae from preeclamptic (n=13) and healthy (n=13) pregnancies were collected and pro-
cessed using the same tissue homogenization-ACN precipitation approach. The 146 markers which
were differentially expressed in the explant models were studied in frozen preeclamptic and normal
147
placental tissues and their statistical significance was calculated by comparing with the controls.
Using the very strictest statistical criteria, 4 of the 146 biomarkers which showed a differentially
increased abundance in the stressed explants in the previous study (Chapter 3) were also elevated
in PE placenta. As described in chapter 3, 1 of these changes were due to hypoxic environment, 2
with TNFα and the remaining 1 as a consequence of oxidative stress (hypoxia-reoxygenation).
Two of the four differentially expressed species were chemically characterized as a phospholipid
(m/z 476.24) and an acylcarnitine (426.35).
5.2 Limitations of current research
5.2.1 Limitations: Chapter 2
Placenta is an organ known to exhibit a high degree of anatomical compartmentalization.
Such variation may affect our results to a significant degree. In my first study where I was inves-
tigating regional differences in healthy placentas, I was very cautious to sample tissue from the
same location every time a specimen was collected to avoid any kind of variation in sampling. For
placental proteomics, generally blood is washed off the tissue, but for our initial experiments we
did not remove external blood. Nevertheless, we found differences that were not due to maternal
blood being present on the exterior of the tissue. We were still able to analyze and find chorionic
plate- or basal plate- specific placental components. Additionally, we studied only two regions of
placenta, investigating a third and potentially an important one: the intervillous region would pro-
vide more information in terms of cellular composition across the placenta. Also, specimens were
collected from a total of 12 placentas which relatively is a small set. A larger set is always preferred
if one is to perform an un-targeted ‘omics’ study. Twelve of sixteen significantly different markers
148
were identified using tandem mass spectrometry, while due to low-abundance of some of them
and high charge state of others we were unable to characterize them using CID alone. In such
situations other ion activation approaches such as HCD/ETD are more useful, which are not com-
patible with QTOF MS.
5.2.2 Limitations: Chapter 3
In study 2, I used placental explants to introduce proposed abnormalities to simulate PE. The
biggest limitation in explant culture there is the need for fresh tissue each time. Processing needs
to start shortly after delivery to prevent loss in tissue viability. To overcome this limitation, we
collected placenta within 3–5 minutes of delivery and started explant culture within 30 minutes of
collection. Another limitation of explant culture is the inefficient transfer of nutrients and oxygen
from the culture medium to the tissue. To make sure that this transfer occurs as efficiently as
possible, our tissue explants are cut into thin slices. We tried to be as consistent as possible with
all other parameters, like weight of the tissue, time required for processing, etc. One of the potential
pitfalls of explant culture is the inability for exact simulation of in vivo conditions while working
in vitro. Explant culture, though, simulates the in vivo conditions more than isolated cell culture
does, because cells in explant culture are in their native environment with cell-to-cell communica-
tion preserved. Additionally, stable culture of placental cells involves cells derived from chori-
ocarcinomas, which may provide a poor reflection of normal or even preelcamptic tissue. This
makes explant culture the superior option since no model can yet imitate the exact in vivo condi-
tions. Additionally, the experiment was designed to incubate placental explant in their respective
environment for a period of 48 hours. Though a lot of changes take place in this period of time,
PE may actually be a consequence of exposure of provocative agents for an extended period of
149
time. Conversely, some changes may happen on a much shorter time scale and be lost by 48 hr. In
addition, I was able to establish identities of only 45% of the biomarkers which were statistically
significant. I believe performing different techniques such as ETD, HCD as available on trap MS
instruments will provide complementary information in terms MS/MS data and thus potentially
help in identification of markers.
5.2.3 Limitations: Chapter 4
As mentioned, placenta exhibits substantial anatomical variations. It is often challenging to
obtain meaningful data in such a case. For our third study which involved studying actual
preeclamptic placentas, the samples came from Erlanger Hospital, University of Tennessee, Chat-
tanooga. We had instructed them to be as consistent as possible in sample site selection while
dissecting tissue from the placentas. Also, the preeclamptic and control placentas were collected
from C-sections as well as vaginal deliveries, from women belonging to different races which adds
a potential source of variation. Due to PE, some women had to undergo early delivery which re-
sulted in earlier gestational age to PE group as compared to uncomplicated pregnancies. A fraction
of these placentas had late term PE. Another major limitation in this study is that the tissue was
frozen immediately after its collection and transported on dry ice from Tennessee to BYU. This
should minimize changes in the placental proteome. However, the effects of freezing in this way
are unknown. The normal placental tissue to be used for comparison was collected at the same site
and preserved and shipped in the same fashion as the preeclamptic tissue. Therefore, the effects of
time and shipping on sample quality will be nullified, and there should be the same loss of speci-
men integrity that might occur. In this study we obtained 13 preeclamptic placentae and equal
number of controls, which is a relatively small set of samples. In addition, we observed only a
150
small fraction of the targeted markers that continued to be different. With the most stringent con-
ditions applied ,only 4 out 146 biomarkers responding to PE stressors changed in PE placenta. One
reason to explain this effect is that changes in response to proposed PE abnormalities may happen
at an early stage in the pregnancy during the events after placentation, and by the time the disease
has been progressed to a full blown PE pregnancy, these changes in the expression of signature
markers might not continue.
Another source of variation which is inevitable arises due to instrument parameters. LC/MS
is sensitive to small, daily changes (for example, temperature). Such changes can create compli-
cations while analyzing mass spectral data, thus giving false results. To minimize instrument var-
iation we used internal reference peaks to normalize all our data.
5.3 Future objectives
5.3.1 Future research objective: Chapter 3
In study 2 (Chapter 3), in total 146 molecules were significantly different between the
stressed explants and controls. 45% of these were identified by performing tandem mass spectrom-
etry. Going forward an attempt can be made to identify the remaining biomolecules by perhaps
using different modes of ion activation such as high energy collision induced dissociation (HCD),
electron transport dissociation (ETD) and electron capture dissociation (ECD). ETD and ECD are
known to aid in identification of peptides with higher charge states as well as keeping intact labile
post-translational modifications during the process of fragmentation, thus helping in the identifi-
cation of modified proteins.
151
Several mediators other than hypoxia, oxidative stress and an accumulation of cytokines in
placenta have been proposed to be involved in the pathogenesis of PE. These factor may act alone
or in combination to produce the symptoms of PE. It would be interesting to see the effect of each
of these mediators on human placental explants and the changes they bring in the LMW biomole-
cules. Thus, as a part of ongoing research to study PE, more stressed conditions or provocative
agents can be applied or introduced to healthy placental explants. Examples include introduction
of sFlt-1, endothelin etc.
5.3.2 Future research objectives: Chapter 4
While studying placentae from preeclamptic pregnancies, we only focused on the 146 mol-
ecules that were significantly altered as a consequence of abnormal conditions introduced in the
healthy placental explants. For a better understanding and a more thorough analysis of PE placenta,
a global ‘omics’ study can be done of all the peaks observed in the MS data to find additional
biomolecules with significant differences between preeclamptic and normal placenta. Analysis and
identification of these species can help throw more light into the mechanism in the fully established
PE.
In our first study to identify regional differences in placenta, we concluded that there exist
subtle differences in biomolecules between chorionic plate and basal plate. Thus it would be inter-
esting to study chorionic plate and basal plate from preeclamptic placentae and compare them to
the two regions from healthy controls. This can provide additional information on the mechanisms
and progression of this disease.
In addition to the peptidomics study, for all of the three projects, a global lipidomic ap-
proach can be performed too, to study multiple classes of lipids using positive ion mode as well
152
as negative ion mode of ionization. A liquid-liquid extraction approach can be used such as
Bligh-Dyer extraction to extract polar as well as non-polar lipids. These can then be injected in
MS directly or subjected to a pre-fractionation step using a suitable chromatography column. Ei-
ther approach will be useful in defining the altered lipidome in preeclamptic conditions as well in
preeclamptic placenta.
153
5.4 References
1. Alvarez, M. T.; Shah, D. J.; Thulin, C. D.; Graves, S. W., Tissue proteomics of the low-
molecular weight proteome using an integrated cLC-ESI-QTOFMS approach. Proteomics
2013, 13 (9), 1400-11.